Why Every Startup Needs an Inverted Demand Generation Funnel, Part III

In part I of this three-part series I introduced the idea of an inverted funnel whereby marketing can derive a required demand generation budget using the sales target and historical conversion rates.  In order to focus on the funnel itself, I made the simplifying assumption that the company’s new ARR target was constant each quarter. 

In part II, I made things more realistic both by quarterizing the model (with increasing quarterly targets) and accounting for the phase lag between opportunity generation and closing that’s more commonly known as “the sales cycle.”  We modeled that phase lag using the average sales cycle length.  For example, if your average sales cycle is 90 days, then opportunities generated in 1Q19 will be modeled  as closing in 2Q19 [1].

There are two things I dislike about this approach:

  • Using the average sales cycle loses information contained in the underlying distribution.  While deals on average may close in 90 days, some deals close in 30 while others may close in 180. 
  • Focusing only on the average often leads marketing to a sense of helplessness. I can’t count the number of times I have heard, “well, it’s week 2 and the pipeline’s light but with a 90-day sales cycle there is nothing we can do to help.”  That’s wrong.  Some deals close more quickly than others (e.g., upsell) so what can we do to find more of them, fast [2].

As a reminder, time-based close rates come from doing a cohort analysis where we take opportunities created in a given quarter and then track not only what percentage of them eventually close, but when they close, by quarter after their creation. 

This allows us to calculate average close rates for opportunities in different periods (e.g., in-quarter, in 2 quarters, or cumulative within 3 quarters) as well an overall (in this case, six-quarter) close rate, i.e., the cumulative sum.  In this example, you can see an overall close rate of 18.7% meaning that, on average, within 6 quarters we close 18.7% of the opportunities that sales accepts.  This is well within what I consider the standard range of 15 to 22%.

Previously, I argued this technique can be quite useful for forecasting; it can also be quite useful in planning.  At the risk of over-engineering, let’s use the concept of time-based close rates  to build an inverted funnel for our 2020 marketing demand generation plan.

To walk through the model, we start with our sales targets and average sales price (ASP) assumptions in order to calculate how many closed opportunities we will need per quarter.  We then drop to the opportunity sourcing section where we use historical opportunity generation and historical time-based close rates to estimate how many closed opportunities we can expect from the existing (and aging) pipeline that we have already generated.  Then we can plug our opportunity generation targets from our demand generation plan into the model (i.e., the orange cells).  The model then calculates a surplus or (gap) between the number of closed opportunities we need and those the model predicts. 

I didn’t do it in the spreadsheet, but to turn that opportunity creation gap into ARR dollars just multiply by the ASP.  For example, in 2Q20 this model says we are 1.1 opportunities short, and thus we’d forecast coming in $137.5K (1.1 * $125K) short of the new ARR plan number.  This helps you figure out if you have the right opportunity generation plan, not just overall, but with respect to timing and historical close rates.

When you discover a gap there are lots of ways to fix it.  For example, in the above model, while we are generating enough opportunities in the early part of the year to largely achieve those targets, we are not generating enough opportunities to support the big uptick in 4Q20.  The model shows us coming in 10.8 opportunities short in 4Q20 – i.e., anticipating a new ARR shortfall of more than $1.3M.  That’s not good enough.  In order to achieve the 4Q20 target we are going to need to generate more opportunities earlier in the year.

I played with the drivers above to do just that, generating an extra 275 opportunities across the year generating surpluses in 1Q20 and 3Q20 that more than offset the small gaps in 2Q20 and 4Q20.  If everything happened exactly according to the model we’d get ahead of plan and 1Q20 and 3Q20 and then fall back to it in 2Q20 and 4Q20 though, in reality, the company would likely backlog deals in some way [3] if it found itself ahead of plan nearing the end of one quarter with a slightly light pipeline the next. 

In concluding this three-part series, I should be clear that while I often refer to “the funnel” as if it’s the only one in the company, most companies don’t have just one inverted funnel.   The VP of Americas marketing will be building and managing one funnel that may look quite different from the VP of EMEA marketing.  Within the Americas, the VP may need to break sales into two funnels:  one for inside/corporate sales (with faster cycles and smaller ASPs) and one for field sales with slower sales cycles, higher ASPS, and often higher close rates.  In large companies, General Managers of product lines (e.g., the Service Cloud GM at Salesforce) will need to manage their own product-specific inverted funnel that cuts across geographies and channels. There’s a funnel for every key sales target in a company and they need to manage them all.

You can download the spreadsheet used in this post, here.

Notes

[1] Most would argue there are two phase lags: the one from new lead to opportunity and the one from opportunity (SQL) creation to close. The latter is the sales cycle.

[2] As another example, inside sales deals tend to close faster than field sales deals.

[3] Doing this could range from taking (e.g., co-signing) the deal one day late to, if policy allows, refusing to accept the order to, if policy enables, taking payment terms that require pushing the deal one quarter back.  The only thing you don’t want to is to have the customer fail to sign the contract because you never know if your sponsor quits (or gets fired) on the first day of the next quarter.  If a deal is on the table, take it.  Work with sales and finance management to figure out how to book it.

The Evolution of Software Marketing: Hey Marketing, Go Get [This]!

As loyal readers know, I’m a reductionist, always trying to find the shortest, simplest way of saying things even if some degree of precision gets lost in the process and even if things end up more subtle than they initially appear.

For example, my marketing mission statement of “makes sales easier” is sometimes misinterpreted as relegating marketing to a purely tactical role, when it actually encompasses far more than that.  Yes, marketing can make sales easier through tactical means like lead generation and sales support, but marketing can also makes sales easier through more leveraged means such as competitive analysis and sales enablement or even more leveraged means such as influencer relations and solutions development or the most leveraged means of picking which markets the company competes in and (with product management) designing products to be easily salable within them.

“Make sales easier” does not just mean lead generation and tactical sales support.

So, in this reductionist spirit, I thought I’d do a historical review of the evolution of enterprise software marketing by looking at its top objective during the thirty-odd years (or should I say thirty odd years) of my career, cast through a fill-in-the-blank lens of, “Hey Marketing, go get [this].”

Hey Marketing, Go Get Leads

In the old days, leads were the focus.  They were tracked on paper and the goal was a big a pile as possible.  These were the days of tradeshow models and free beer:  do anything to get people come by the booth – regardless of whether they have any interest in or ability to buy the software.  Students, consultants, who cares?  Run their card and throw them in the pile.  We’ll celebrate the depth of the pile at the end of the show.

Hey Marketing, Go Get Qualified Leads

Then somebody figured out that all those students and consultants and self-employed people who worked at companies way outside the company’s target customer size range and couldn’t actually buy our software.  So the focus changed to get qualified leads.  Qualified first basically meant not unqualified:

  • It couldn’t be garbage, illegible, or duplicate
  • It couldn’t be self-employed, students, or consultants
  • It couldn’t be other people who clearly can’t buy the software (e.g., in the wrong country, at too small a company, in a non-applicable industry)

Then people realized that not all not-unqualified leads were the same. 

Enter lead scoring.  The first systems were manual and arbitrarily defined:  e.g., let’s give 10 points for target companies, 10 points for a VP title, and 15 points if they checked buying-within-6-months on the lead form.  Later systems got considerably more sophisticated adding both firmographic and behavioral criteria (e.g., downloaded the Evaluation Guide).  They’d even have decay functions where downloading a white paper got you 10 points, but you’d lose a point every week since if there you had no further activity. 

The problem was, of course, that no one ever did any regressions to see if A leads actually were more likely to close than B leads and so on.  At one company I ran, our single largest customer was initially scored a D lead because the contact downloaded a white paper using his Yahoo email address.  Given such stories and a general lack of faith in the scoring system, operationally nobody ever treated an A lead differently from a D lead – they’d all get “6×6’ed” (6 emails and 6 calls) anyway by the sales development reps (SDRs).  If the score didn’t differentiate the likelihood of closing and the SDR process was score-invariant, what good was scoring? The answer: not much.

Hey Marketing, Go Get Pipeline

Since it was seemingly too hard to figure out what a qualified lead was, the emphasis shifted.  Instead of “go get leads” it became, “go get pipeline.”  After all, regardless of score, the only leads we care about are those that turn into pipeline.  So, go get that.

Marketing shifted emphasis from leads to pipeline as salesforce automation (SFA) systems were increasingly in place that made pipeline easier to track.  The problem was that nobody put really good gates on what it took to get into the pipeline.  Worse yet, incentives backfired as SDRs, who were at the time almost always mapped directly to quota-carrying reps (QCRs), were paid incentives when leads were accepted as opportunities.  “Heck,” thinks the QCR, “I’ll scratch my SDR’s back in order to make sure he/she keeps scratching mine:  I’ll accept a bunch of unqualified opportunities, my SDR will get paid a $200 bonus on each, and in a few months I’ll just mark them no decision.  No harm, no foul. “Except the pipeline ends up full of junk and the 3x self-fulfilling pipeline coverage prophecy is developed.  Unless you have 3x coverage, your sales manager will beat you up, so go get 3x coverage regardless of whether it’s real or not.  So QCRs stuff bad opportunities into the pipeline which in turn converts at a lower rate which in turn increases the coverage goal – i.e., “heck, we’re only converting pipeline at 25%, so now we need 4x coverage!”  And so on.

At one point in my career I actually met a company with 100x pipeline coverage and 1% conversion rates. 

Hey Marketing, Go Get Qualified Opportunities (SQLs)

Enter the sales qualified lead (SQL). Companies realize they need to put real emphasis on someone, somewhere in the process defining what’s real and what not.  That someone ends up the QCR and it’s now their job to qualify opportunities as they are passed over and only accept those that both look real and meet documented criteria.  Management is now focused on SQLs.  SQL-based metrics, such as cost-per-SQL or SQL-to-close-rate, are created and benchmarked.  QCRs can no longer just accept everything and no-decision it later and, in fact, there’s less incentive to anyway as SDRs are no longer basically working for the QCRs, but instead for “the process” and they’re increasingly reporting into marketing to boot.  Yes, SDRs will be paid on SQLs accepted by sales, but sales is going to be held highly accountable for what happens to the SQLs they accept. 

Hey Marketing, Go Get Qualified Opportunities Efficiently

At this point we’ve got marketing focused on SQL generation and we’ve built a metrics-driven inbound SDR team to process all leads. We’ve eliminated the cracks between sales and marketing and, if we’re good, we’ve got metrics and reporting in place such that we can easily see if leads or opportunities are getting stuck in the pipeline. Operationally, we’re tight.

But are we efficient? This is also the era of SaaS metrics and companies are increasingly focused not just on growth, but growth efficiency.  Customer acquisition cost (CAC) becomes a key industry metric which puts pressure on both sales and marketing to improve efficiency.  Sales responds by staffing up sales enablement and sales productivity functions. Marketing responds with attribution as a way to try and measure the relative effectiveness of different campaigns.

Until now, campaign efficiency tended to be measured a last-touch attribution basis. So when marketers tried to calculate the effectiveness of various marketing campaigns, they’d get a list of closed deals, and allocate the resultant sales to campaigns by looking at the last thing someone did before buying. The predictable result: down-funnel campaigns and tools got all of the credit and up-funnel campaigns (e.g., advertising) got none.

People pretty quickly realized this was a flawed way to look at things so, happily, marketers didn’t shoot the propellers off their marketing planes by immediately stopping all top-of-funnel activity. Instead, they kept trying to find better means of attribution.

Attribution systems, like Bizible, came along which tried to capture the full richness of enterprise sales. That meant modeling many different contacts over a long period of time interacting with the company via various mechanisms and campaigns. In some ways attribution became like search: it wasn’t whether you got the one right answer, it was whether search engine A helped you find relevant documents better than search engine B. Right was kind of out the question. I feel the same way about attribution. Some folks feel it doesn’t work at all. My instinct is that there is no “right” answer but with a good attribution system you can do better at assessing relative campaign efficiency than you can with the alternatives (e.g., first- or last-touch attribution).

After all, it’s called the marketing mix for a reason.

Hey Marketing, Go Get Qualified Opportunities That Close

After the quixotic dalliance with campaign efficiency, sales got marketing focused back on what mattered most to them. Sales knew that while the bar for becoming a SQL was now standardized, that not all SQLs that cleared it were created equal. Some SQLs closed bigger, faster, and at higher rates than others. So, hey marketing, figure out which ones those are and go get more like them.

Thus was born the ideal customer profile (ICP). In seed-stage startups the ICP is something the founders imagine — based on the product and target market they have in mind, here’s who we should sell to. In growth-stage startups, say $10M in ARR and up, it’s no longer about vision, it’s about math.

Companies in this size range should have enough data to be able to say “who are our most successful customers” and “what do they have in common.” This involves doing a regression between various attributes of customers (e.g., vertical industry, size, number of employees, related systems, contract size, …) and some success criteria. I’d note that choosing the success criteria to regress against is harder than meetings the eye: when we say we find to find prospects most like our successful customers, how are we defining success?

  • Where we closed a big deal? (But what if it came at really high cost?)
  • Where we closed a deal quickly? (But what if they never implemented?)
  • Where they implemented successfully? (But what if they didn’t renew?)
  • Where they renewed once? (But what if they didn’t renew because of uncontrollable factor such as being acquired?)
  • Where they gave us a high NPS score? (But what if, despite that, they didn’t renew?)

The Devil really is in the detail here. I’ll dig deeper into this and other ICP-related issues one day in a subsequent post. Meantime, TOPO has some great posts that you can read.

Once you determine what an ideal customer looks like, you can then build a target list of them and enter into the world of account-based marketing (ABM).

Hey Marketing, Go Get Opportunities that Turn into Customers Who Renew

While sales may be focused simply on opportunities that close bigger and faster than the rest, what the company actually wants is happy customers (to spread positive word of mouth) who renew. Sales is typically compensated on new orders, but the company builds value by building its ARR base. A $100M ARR company with a CAC ratio of 1.5 and churn rate of 20% needs to spend $30M on sales and marketing just to refill the $20M lost to churn. (I love to multiply dollar-churn by the CAC ratio to figure out the real cost of churn.)

What the company wants is customers who don’t churn, i.e., those that have a high lifetime value (LTV). So marketing should orient its ICP (i.e., define success in terms of) not just likelihood to {close, close big, close fast} but around likelihood to renew, and potentially not just once. Defining different success criteria may well produce a different ICP.

Hey Marketing, Go Get Opportunities that Turn into Customers Who Expand

In the end, the company doesn’t just want customers who renew, even if for a long time. To really the build the value of the ARR base, the company wants customers who (1) are relatively easily won (win rate) and relatively quickly (average sales cycle) sold, (2) who not only renew multiple times, but who (3) expand their contracts over time.

Enter net dollar expansion rate (NDER), the metric that is quickly replacing churn and LTV, particularly with public SaaS companies. In my upcoming SaaStr 2020 talk, Churn is Dead, Love Live Net Dollar Expansion Rate, I’ll go into why this happening and why companies should increasingly focus on this metric when it comes to thinking about the long-term value of their ARR base.

In reality, the ultimate ICP is built around customers who meet the three above criteria: we can sell them fairly easily, they renew, and they expand. That’s what marketing needs to go get!

Kellblog's 10 Predictions for 2020

As I’ve been doing every year since 2014, I thought I’d take some time to write some predictions for 2020, but not without first doing a review of my predictions for 2019.  Lest you take any of these too seriously, I suggest you look at my batting average and disclaimers.

Kellblog 2019 Predictions Review

1.  Fred Wilson is right, Trump will not be president at the end of 2019.  PARTIAL.  He did get impeached after all, but that’s a long way from removed or resigned. 

2.  The Democratic Party will continue to bungle the playing of its relatively simple hand.  HIT.  This is obviously subjective and while I think they got some things right (e.g., delaying impeachment), they got others quite wrong (e.g., Mueller Report messaging), and continue to play more left than center which I believe is a mistake.

3.  2019 will be a rough year for the financial markets.  MISS.  The Dow was up 22% and the NASDAQ was up 35%.  Financially, maybe the only thing that didn’t work in 2019 were over-hyped IPOs.  Note to self:  avoid quantitative predictions if you don’t want to risk ending up very wrong.  I am a big believer in regression to the mean, but nailing timing is the critical (and virtually impossible) part.  Nevertheless, I do use tables like these to try and eyeball situations where it seems a correction is needed.  Take your own crack at it.

4.  VC tightens.  MISS.  Instead of tightening, VC financing hit a new record.  The interesting question here is whether mean reversion is relevant.  I’d argue it’s not – the markets have changed structurally such that companies are staying private far longer and thus living off venture capital (and/or growth-stage private equity) in ways not previously seen.  Mark Suster did a great presentation on this, Is VC Still a Thing, where he explains these and other changes in VC.  A must read.

5. Social media companies get regulated.  PARTIAL.  While “history may tell us the social media regulation is inevitable,” it didn’t happen in 2019.  However, the movement continued to gather steam with many Democratic presidential candidates calling for reform and, more notably, none other than Facebook investor Roger McNamee launching his attack on social media via his book Zucked: Waking Up To The Facebook Catastrophe.  As McNamee says, “it’s an issue of ‘right vs. wrong,’ not ‘right vs. left.’”

 

6. Ethics make a comeback.  HIT.  Ethics have certainly been more discussed than ever and related to the two reasons I cited:  the current administration and artificial intelligence.  The former forces ethics into the spotlight on a daily basis; the later provokes a slew of interesting questions, from questions of accidental bias to the trolley car problem.  Business schools continue to increase emphasis on ethics.  Mark Benioff has led a personal crusade calling for what he calls a new capitalism.

7.  Blockchain, as an enterprise technology, fades away.  HIT.  While I hate to my find myself on the other side of Ray Wang, I’m personally not seeing much traction for blockchain in the enterprise.  Maybe I’m running with the wrong crowd.  I have always felt that blockchain was designed for one purpose (to support cybercurrency), hijacked to another, and ergo became a vendor-led technology in search of a business problem.  McKinsey has a written a sort of pre-obituary, Blockchain’s Occam Problem, which was McKinsey Quarterly’s second most-read article of the year.  The 2019 Blockchain Opportunity Summit’s theme was Is Blockchain Dead?  No. Industry Experts Join Together to Share How We Might Not be Using it Right which also seems to support my argument. 

8.  Oracle enters decline phase and is increasingly seen as a legacy vendor.  HIT.  Again, this is highly subjective and some people probably concluded it years ago.  My favorite support point comes from a recent financial analyst note:  “we believe Oracle can sustain ~2% constant currency revenue growth, but we are dubious that Oracle can improve revenue growth rates.”  That pretty much says it all.

9.  ServiceNow and/or Splunk get acquired.  MISS.  While they’re both great businesses and attractive targets, they are both so expensive only a few could make the move – and no one did.  Today, Splunk is worth $24B and ServiceNow a whopping $55B.

10.  Workday succeeds with its Adaptive Insights agenda.  HIT.  Changing general ledgers is a heart transplant while changing planning systems is a knee replacement.  By acquiring Adaptive, Workday gave itself another option – and a far easier entry point – to get into corporate finance departments.  While most everyone I knew scratched their head at the enterprise-focused Workday acquiring a more SMB-focused Adaptive, Workday has done a good job simultaneously leaving Adaptive alone-enough to not disturb its core business while working to get the technology more enterprise-ready for its customers.  Whether that continues I don’t know, but for the first 18 months at least, they haven’t blown it.  This remains high visibility to Workday as evidenced by the Adaptive former CEO (and now Workday EVP of Planning) Tom Bogan’s continued attendance on Workday’s quarterly earnings calls.

With the dubious distinction of having charitably self-scored a 6.0 on my 2019 predictions, let’s fearlessly roll out some new predictions for 2020.

Kellblog 2020 Predictions

1.  Ongoing social unrest. The increasingly likely trial in the Senate will be highly contentious, only to be followed by an election that will be highly contentious as well.  Beyond that, one can’t help but wonder if a defeated Trump would even concede, which could lead to a Constitutional Crisis of the next level. Add to all that the possibility of a war with Iran.  Frankly, I am amazed that the Washington, DC continuous distraction machine hasn’t yet materially damaged the economy.  Like many in Silicon Valley, I’d like Washington to quietly go do its job and let the rest of us get back to doing ours.  The reality TV show in Washington is getting old and, happily, I think many folks are starting to lose interest and want to change the channel.

2.  A desire for re-unification.  I remain fundamentally optimistic that your average American – Republican, Democrat, or the completely under-discussed 38% who are Independents — wants to feel part of a unified, not a divided, America.  While politicians often try to leverage the most divisive issues to turn people into single-issue voters, the reality is that far more things unite us as Americans than divide us.  Per this recent Economist/YouGov wide-ranging poll, your average American looks a lot more balanced and reasonable than our political party leaders.  I believe the country is tired of division, wants unification, and will therefore elect someone who will be seen as able to bring people together.  We are stronger together.

3.  Climate change becomes the new moonshot.  NASA’s space missions didn’t just get us to the moon; they produced over 2,000 spin-off technologies that improve our lives every day – from emergency “space” blankets to scratch-resistant lenses to Teflon-coated fabrics.  Instead of seeing climate change as a hopeless threat, I believe in 2020 we will start to reframe it as the great opportunity it presents.  When we mobilize our best and brightest against a problem, we will not only solve it, but we will create scores to hundreds of spin-off technologies that will benefit our everyday lives in the process.  See this article for information on 10 startups fighting climate change, this infographic for an overview of the kinds of technologies that could alleviate it, or this article for a less sanguine view on the commitment required and extent to which we actually can de-carbonize the air. Or check out this startup which makes “trees” that consume the pollution of 275 regular trees.

4.  The strategic chief data officer (CDO).  I’m not a huge believer in throwing an “O” at every problem that comes along, but the CDO role is steadily becoming mainstream – in 2012 just 12% of F1000 companies reported having a CDO; in 2018 that’s up to 68%.  While some of that growth was driven by defensive motivations (e.g., compliance), increasingly I believe that organizations will define the CDO more strategically, more broadly, and holistically as someone who focuses on data, its cleanliness, where to find it, where it came from, its compliance with regulations as to its usage, its value, and how to leverage it for operational and strategic advantage.   These issues are thorny, technical, and often detail-oriented and the CIO is simply too busy with broader concerns (e.g., digital transformation, security, disruption).  Ergo, we need a new generation of chief data officers who want to play both offense and defense, focused not just tactically on compliance and documentation, but strategically on analytics and the creation of business value for the enterprise. This is not a role for the meek; only half of CDOs succeed and their average tenure is 2.4 years.  A recent Gartner CDO study suggests that those who are successful take a more strategic orientation, invest in a more hands-on model of supporting data and analytics, and measure the business value of their work.

5.  The ongoing rise of DevOps.   Just as agile broke down barriers between product management and development so has DevOps broken down walls between development and operations.  The cloud has driven DevOps to become one of the hottest areas of software in recent years with big public company successes (e.g., Atlassian, Splunk), major M&A (e.g., Microsoft acquiring GitHub), and private high-flyers (e.g., HashiCorp, Puppet, CloudBees).  A plethora of tools, from configuration management to testing to automation to integration to deployment to multi-cloud to performance monitoring are required to do DevOps well.  All this should make for a $24B DevOps TAM by 2023 per a recent Cowen & Company report.  Ironically though, each step forward in deployment is often a step backward in developer experience, why is one reason why I decided to work with Kelda in 2019.

6. Database proliferation slows.  While 2014 Turning Award winner Mike Stonebraker was right over a decade ago when he argued in favor of database specialization (One Size Fits All:  An Idea Whose Time Has Come and Gone), I think we may now too much of a good thing.   DB Engines now lists 350 different database systems of 14 different types (e.g., relational, graph, time series, key-value). Crunchbase lists 274 database (and database-related) startups.  I believe the database market is headed for consolidation.  One of the first big indicators of a resurgence in database sanity was the failure of the (Hadoop-based) data lake, which happened in 2018-2019 and was the closest thing I’ve seen to déjà vu in my professional career – it was as if we learned nothing from the Field of Dreams enterprise data warehouse of the 1990s (“build it and they will come”).  Moreover, after a decade of developer-led database selection, developers and now re-realizing what database people knew along – that a lot of the early NoSQL movement was akin to throwing out the ACID transaction baby with the tabular schema bathwater.

7.  A new, data-layer approach to data loss prevention (DLP).  I always thought DLP was a great idea, especially the P for prevention.  After all, who wants tools that can help with forensics after a breach if you could prevent one from happening at all — or at least limit one in progress?  But DLP doesn’t seem to work:  why is it that data breaches always seem to be measured not in rows, but in millions of rows?  For example, Equifax was 143M and Marriott was 500M.  DLP has many known limitations.  It’s perimeter-oriented in a hybrid cloud world of dissolving perimeters and it’s generally offline, scanning file systems and database logs to find “misplaced data.”  Wouldn’t a better approach be to have real-time security monitored and enforced at the data layer, just the same way as it works at the network and application layer?  Then you could use machine learning to understand normal behavior, detect anomalous behavior, and either report it — or stop it — in real time.  I think we’ll see such approaches come to market in 2020, especially as cloud services like Snowflake, RDS, and BigQuery become increasingly critical components of the data layer.

8. AI/ML continue to see success in highly focused applications.  I remain skeptical of vendors with broad claims around “enterprise AI” and remain highly supportive of vendors applying AI/ML to specific problems (e.g., Moveworks and Astound who both provide AI/ML-based trouble-ticket resolution).  In the end, AI and ML are features, not apps, and while both technologies can be used to build smart applications, they are not applications unto themselves.  In terms of specificity, the No Free Lunch Theorem reminds us that any two optimization techniques perform equivalently when averaged across all possible problems – meaning that no one modeling technique can solve everything and thus that AI/ML is going to be about lots of companies applying different techniques to different problems.   Think of AI/ML more as a toolbox than a platform.  There will not be one big winner in enterprise AI as there was in enterprise applications or databases.  Instead, there will be lots of winners each tackling specific problems.  The more interesting battles will those between systems of intelligence (e.g., Moveworks) and systems of record (e.g., ServiceNow) with the systems-of-intelligence vendors running Trojan Horse strategies against systems-of-record vendors (first complementing but eventually replacing them) while the system-of-record vendors try to either build or acquire systems of intelligence alongside their current offerings. 

9.  Series A rounds remain hard.  I think many founders are surprised by the difficulty of raising A rounds these days.  Here’s the problem in a nutshell:

  • Seed capital is readily available via pre-seed and seed-stage investments from angel investors, traditional early-stage VCs, and increasingly, seed funds.  Simply put, it’s not that hard to raise seed money.
  • Companies are staying in the seed stage longer (a median of 1.6 years), increasingly extending seed rounds, and ergo raising more money during seed stage (e.g., $2M to $4M).
  • Such that, companies are now expected to really have achieved something in order to raise a Series A.  After all, if you have been working for 2 years and spent $3M you better have an MVP product, a handful of early customers, and some ARR to show for it – not just a slide deck talking about a great opportunity.

Moreover, you should be making progress roughly in line with what you said at the outset and, if you took seed capital from a traditional VC, then they better be prepared to lead your round otherwise you will face signaling risk that could imperil your Series A.

Simply put, Series A is the new chokepoint.  Or, as Suster likes to say, the Series A and B funnel hasn’t really changed – we’ve just inserted a new seed funnel atop it that is 3 times larger than it used to be.

10.  Autonomy’s former CEO gets extradited.  Silicon Valley is generally not a place of long memories, but I saw the unusual news last month that the US government is trying to extradite Autonomy founder and former CEO Mike Lynch from the UK to face charges.  You might recall that HP, in the brief era under Leo Apotheker, acquired enterprise search vendor Autonomy in August, 2011 for a whopping $11B only to write off about $8.8B under subsequent CEO Meg Whitman a little more than a year later in November, 2012.  Computerworld provides a timeline of the saga here, including a subsequent PR war, US Department of Justice probe, UK Serious Fraud Office investigation (later dropped), shareholder lawsuits, proposed settlements, more lawsuits including Lynch’s suing HP for $150M for reputation damages, and HP’s spinning-off the Autonomy assets.  Subsequent to Computerworld’s timeline, this past May Autonomy’s former CFO was sentenced to five years in prison.  This past March, the US added criminal charges of securities fraud, wire fraud, and conspiracy against Lynch.  Lynch continues to deny all wrongdoing, blames the failed acquisition on HP, and even maintains a website to present his point of view on the issues.  I don’t have any special legal knowledge or specific knowledge of this case, but I do believe that if the US government is still fighting this case, still adding charges, and now seeking extradition, that they aren’t going to give up lightly, so my hunch is that Lynch does come to the US and face these charges. 

More broadly, regardless of how this particular case works out, in a place so prone to excess, where so much money can be made so quickly, frauds will periodically happen and it’s probably the most under-reported class of story in Silicon Valley.  Even this potentially huge headline case – the proposed extradition of a British billionaire tech mogul —  never seems to make page one news.  Hey, let’s talk about something positive like Loft’s $175M Series C instead.

To finish this up, I’ll add a bonus prediction:  Dave doesn’t get a traditional job in 2020.  While I continue to look at VC-backed startup and/or PE-backed CEO opportunities, I am quite enjoying my work doing a mix of boards, advisory relationships, and consulting gigs.  While I remain interested in looking at great CEO opportunities, I am also interested in adding a few more boards to my roster, working on stimulating consulting projects, and a few more advisory relationships as well.

I wish everyone a happy, healthy, and above-plan 2020.

Why Every Startup Needs an Inverted Demand Generation Funnel, Part II

In the previous post, I introduced the idea of an inverted demand generation (demandgen) funnel which we can use to calculate a marketing demandgen budget based given a sales target, an average sales price (ASP), and a set of conversion rates along the funnel. This is a handy tool, isn’t hard to make, and will force you into the very good habit of measuring (and presumably improving) a set of conversion rates along your demand funnel.

In the previous post, as a simplifying assumption, we assumed a steady-state situation where a company had a $2M new ARR target every quarter. The steady-state assumption allowed us to ignore two very real factors that we are going to address today:

  • Time. There are two phase-lags along the funnel. MQLs might take a quarter to turn into SALs and SALs might take two quarters to turn into closed deals. So any MQL we generate now won’t likely become a closed deal until 3 quarters from now.
  • Growth. No SaaS company wants to operate at steady state; sales targets go up every year. Thus if we generate only enough MQLs to hit this-quarter’s target we will invariably come up short because those MQLs are working to support a (presumably larger) target 3 quarters in the future.

In order to solve these problems we will start with the inverted funnel model from the previous post and do three things:

  • Quarter-ize it. Instead of just showing one steady-state quarter (or a single year), we are going to stretch the model out across quarters.
  • Phase shift it. If SALs take two quarters to close and MQLs take 1 quarter to become SALS we will reflect this in the model, by saying 4Q20 deals need come from SALs generated in 2Q20 which in turn come from MQLs generated in 1Q20.
  • Extend it. Because of the three-quarter phase shift, the vast majority of the MQLs we’ll be generating 2020 are actually to support 2021 business, so we need to extend the model in 2021 (with a growth assumption) in order to determine how big of a business we need to support.

Here’s what the model looks like when you do this:

You can see that this model generates a varying demandgen budget based on the future sales targets and if you play with the drivers, you can see the impact of growth. At 50% new ARR growth, we need a $1.47M demandgen budget in 2020, at 0% we’d need $1.09M, and at 100% we’d need $1.85M.

Rather than walk through the phase-shifting with words, let me activate Excel’s trace-precedents feature so you can see how things flow:

With these corrections, we have transformed the inverted funnel into a pretty realistic tool for modeling MQL requirements of the company’s future growth plan.

Other Considerations

In reality, your business may consist of multiple funnels with different assumption sets.

  • Partner-sourced deals are likely to have smaller deal sizes (due to margin given to the channel) but faster conversion timeframes and higher conversion rates. (Because we will learn about deals later in the cycle, hear only about the good ones, and the partner may expedite the evaluation process.)
  • Upsell business will almost certainly have smaller deal sizes, faster conversion timeframes, and much higher conversion rates than business to entirely new customers.
  • Corporate (or inside) sales is likely to have a materially different funnel from enterprise sales. Using a single funnel that averages the two might work, provided your mix isn’t changing, but it is likely to leave corporate sales starving for opportunities (since they do much smaller deals, they need many more opportunities).

How many of these funnels you need is up to you. Because the model is particularly sensitive to deal size (given a constant set of conversion rates) I would say that if a certain type of business has a very different ASP from the main business, then it likely needs its own funnel. So instead of building one funnel that averages everything across your company, you might be three — e.g.,

  • A new business funnel
  • An upsell funnel
  • A channel funnel

In part III of this series, we’ll discuss how to combine the idea of the inverted funnel with time-based close rates to create an even more accurate model of your demand funnel.

The spreadsheet I made for this series of posts is available here.

Why Every Startup Needs an Inverted Demand Generation Funnel, Part I

Does my company spend too much on marketing? Too little? How I do know? What is the right level of marketing spend at an enterprise software startup? I get asked these questions all the time by startup CEOs, CMOs, marketing VPs, and marketing directors.

You can turn to financial benchmarks, like the KeyBanc Annual SaaS Survey for some great high-level answers. You can subscribe to SiriusDecisions for best practices and survey data. Or you can buy detailed benchmark data [1] from OPEXEngine. These are all great sources and I recommend them heartily to anyone who can afford them.

But, in addition to sometimes being too high-level [2], there is one key problem with all these forms of benchmark data: they’re not about you. They’re not based on your operating history. While I certainly recommend that executives know their relevant financial benchmarks, there’s a difference between knowing what’s typical for the industry and what’s typical for you.

So, if you want to know if your company is spending enough on marketing [3], the first thing you should do is to make an inverted demand generation (aka, demandgen) funnel to figure out if you’re spending enough on demandgen. It’s quite simple and I’m frankly surprised how few folks take the time to do it.

Here’s an inverted demandgen funnel in its simplest form:

Inverted demandgen funnel

Let’s walk through the model. Note that all orange cells are drivers (inputs) and the white cells are calculations (outputs). This model assumes a steady-state situation [4] where the company’s new ARR target is $2,000,000 each quarter. From there, we simply walk up the funnel using historical deal sizes and conversion rates [5].

  • With an average sales price (ASP) of $75,000, the company needs to close 27 opportunities each quarter.
  • With a 20% sales qualified lead (SQL) to close rate we will need 133 SQLs per quarter.
  • If marketing is responsible for generating 80% of the sales pipeline, then marketing will need to generate 107 of those SQLs.
  • If our sales development representatives (SDRs) can output 2.5 opportunities per week then we will need 5 SDRs (rounding up).
  • With an 80% SAL to SQL conversion rate we will need 133 SALs per quarter.
  • With a 10% MQL to SAL conversion rate we will need 1,333 MQLs per quarter.
  • With a cost of $250 per MQL, we will need a demandgen budget [6] of $333,333 per quarter.

The world’s simplest way to calculate the overall marketing budget at this point would be to annualize demandgen to $1.3M and then double it, assuming the traditional 50/50 people/programs ratio [7].

Not accounting for phase lag or growth (which will be the subjects of part II and part III of this post), let’s improve our inverted funnel by adding benchmark and historical data.

Let’s look at what’s changed. I’ve added two columns, one with 2019 actuals and one with benchmark data from our favorite source. I’ve left the $2M target in both columns because I want to compare funnels to see what it would take to generate $2M using either last year’s or our benchmark’s conversion rates. Because I didn’t want to change the orange indicators (of driver cells) in the left column, when we have deviations from the benchmark I color-coded the benchmark column instead. While our projected 20% SQL-to-close rate is an improvement from the 18% rate in 2019, we are still well below the benchmark figure of 25% — hence I coded the benchmark red to indicate a problem in this row. Our 10% MQL-to-SQL conversion rate in the 2020 budget is a little below the benchmark figure of 12%, so I coded it yellow. Our $250 cost/MQL is well below the benchmark figure of $325 so I coded it green.

Finally, I added a row to show the relative efficiency improvement of the proposed 2020 budget compared to last year’s actuals and the benchmark. This is critical — this is the proof that marketing is raising the bar on itself and committed to efficiency improvement in the coming year. While our proposed funnel is overall 13% more efficient than the 2019 funnel, we still have work to do over the next few years because we are 23% less efficient than we would be if we were at the benchmark on all rates.

However, because we can’t count on fixing everything at once, we are taking a conservative approach where we show material improvement over last year’s actuals, but not overnight convergence to the benchmark — which could take us from kaizen-land to fantasy-land and result in a critical pipeline shortage downstream.

Moreover because this approach shows not only a 13% overall efficiency improvement but precisely where you expect it to come from, the CEO can challenge sales and marketing leadership:

  • Why are we expecting to increase our ASP by $5K to $75K?
  • Why do you think we can improve the SQL-to-close rate from 18% to 20% — and what you are doing to drive that improvement? [8]
  • What are we doing to improve the MQL-to-SAL conversion rate?
  • How are we going to improve our already excellent cost per MQL by $25?

In part II and part III of this post, we’ll discuss two ways of modeling phase-lag, modeling growth, and the separation of the new business and upsell funnels.

You can download my spreadsheet for this post, here.

Notes

[1] For marketing or virtually anything else.

[2] i.e., looking at either S&M aggregated or even marketing overall.

[3] The other two pillars of marketing are product marketing and communications. The high-level benchmarks can help you analyze spend on these two areas by subtracting your calculated demandgen budget from the total marketing budget suggested by a benchmark to see “what’s left” for the other two pillars. Caution: sometimes that result is negative!

[4] The astute reader will instantly see two problems: (a) phase-lag introduced by both the lead maturation (name to MQL) and sales (SQL to close) cycles and (b) growth. That is, in a normal high-growth startup, you need enough leads not to generate this quarter’s new ARR target but the target 3-4 quarters out, which is likely to be significantly larger. Assuming a steady-state situation gets rid of both these problems and simplifies the model. See part II and part III of this post for how I like to manage that added real-world complexity.

[5] Hint: if you’re not tracking these rates, the first good thing about this model is that it will force you to do so.

[6] When I say demandgen budget, I mean money spent on generating leads through marketing campaigns. Sometimes that very directly (e.g., adwords). Other times it’s a bit indirectly (e.g., an SEO program). I do not include demandgen staff because I am trying to calculate the marginal cost of generating an extra MQL. That is, I’m not trying to calculate what the company spends, in total, on demandgen activities (which would include salary, benefits, stock-based comp, etc. for demandgen staff) but instead the marketing programs cost to generate a lead (e.g., in case we need to figure out how much to budget to generate 200 more of them).

[7] In an increasingly tech-heavy world where marketing needs to invest a lot in infrastructure as well, I have adapted the traditional 50/50 people/programs rule to a more modern 45/45/10 people/programs/infrastructure rule, or even an infrastructure-heavy split of 40/40/20.

[8] Better closing tools, an ROI calculator, or a new sales training program could all be valid explanations for assuming an improved close rate.

Why I'm Advising Kelda

A few months ago I signed up to be an advisor to Kelda, and I thought I’d do a quick post to talk about the company and why I decided to sign up.

What is Kelda?

Kelda provides developer sandboxes in a customer’s cloud within their Kubernetes cluster. Why does this matter?

  • The world is moving to cloud computing at a rapid place.
  • Cloud computing is moving away from virtual machines as the unit of abstraction and towards containers, microservices, and serverless architectures.
  • The exact technologies that make microservices powerful in production environments have made the development experience worse.

In short, nobody was thinking much about developers when they started migrating to these new architectures.

Think for a minute about being a developer building a microservices-based application. Then think about testing it. Your code has dependencies on scores or hundreds of microservices which in turn have dependencies on other microservices. Any or all of these microservices are themselves changing over time. How you are you supposed to find a stable test-bed on which to test your code?

Unlike production environments, run by DevOps teams with a sophisticated CI/CD platform, development environments are often primitive by comparison. Tools for collecting dependencies are not robust. Developers often have to test on their own laptops, running all the required microservices locally, which elongates test cycles because of slow performance. Moreover, debugging is potentially complicated by non-deterministic interactions among microservices.

Kelda solves all that by effectively spinning up a private, stable, server-based Kubernetes cluster where developers can test their code. If that sounds pretty practical, well it is. If that sounds pedestrian, remember that one of VMware’s top early use-case was … stable test environments for QA teams across different version of operating systems, middleware, and databases. Pragmatic solutions often generalize way beyond their initial landing point.

For more technical information on Kelda, here’s a link where you can download their white paper. And here’s an excerpt that sums things up quite nicely:

Why Did I Sign Up to Advise Kelda?

There are always many reasons behind such a decision, so in no particular order:

  • The awesome founder, Ethan Jackson, who put his Berkeley computer science PhD on the back burner in order create the company. I like that this isn’t his first corporate rodeo (he worked at Nicira –> VMware) for 5 years. I also like the burn-the-ships level of commitment.
  • The practical logic behind the product idea. Remember the famous William Gibson quote: “the future is already here — it’s just not very evenly distributed.” When you’re working at the cutting edge, the next step looks kind of obvious. So while this looks very high-tech to me, it looks pretty obvious to Ethan and, in my humble opinion, a lot of people have been very successful doing the next pretty-obvious thing (e.g., from PeopleSoft building apps atop Oracle to NetSuite taking financials to the cloud to Palo Alto Networks doing application-based firewalls).
  • The trends driving the company. Kelda is dead center of the movement to containers and microservices-based architectures in the cloud. The technology elite can use all these technologies today. Kelda makes them more accessible to the typical corporate development shop.

Should SDRs Report to Sales or Marketing?

Slowly and steadily, over the past decade, the industry has evolved from a mentality of “all salesreps must do everything” – including some percent of their time prospecting — to one of specialization.  We, with the help of books like Predictable Revenue, have collectively decided that in-bound lead processing is different from outbound lead prospecting is different from low-end, velocity sales is different from high-end, enterprise sales.

Despite the old-school, almost-character-building emphasis on prospecting, we have collectively realized that having our top hunters dialing for dollars and digging through inbound leads isn’t, well, the best use of their time.

Industrialization typically involves specialization and the industrialization of once purely artisanal software sales has been no exception.  As part of this specialization the sales development representative (SDR) role has risen to prominence.  In this post, we’ll do a quick review of what SDRs typically do and discuss the relative merits of having them report into sales vs. marketing.

“Everyone under 25 in San Francisco is an SDR.” – Anonymous startup CEO

SDRs Bridge the Two Departments

SDRs typically form the bridge between sales and marketing.  A typical SDR job is take inbound leads from marketing, perform some basic BANT-style [1] qualification on them, and then pass them to sales if indicated. While SDRs typically have activity quotas (e.g., 50 calls/day) they should be primarily measured on the number of opportunities they create per week. In enterprise software, typically that quota is 2-3 oppties/week. 

As companies get bigger they tend to separate SDRs into two groups:

  • Inbound SDRs, those who only process in-bound leads, and
  • Outbound SDRs, those who primarily do targeted outreach over the phone or email

Being an SDR is a hard job.  Typical SDR challenges include:

  • Adhering to service-level agreements for all leads (i.e., touches with timeframes)
  • Contacting prospects in an increasingly spam-hostile, call-hostile environment
  • Figuring out which leads to work on the hardest (e.g., which merit homework to customize the message and which don’t)
  • Remembering that their job is to sell meetings and not product [2]
  • Supporting multiple salespeople with often conflicting priorities [3]
  • Managing the conflict between supporting salespeople and executing the process
  • Getting salespeople to show-up at the hand-off meeting [4]
  • Avoiding burnout in a high-pressure environment

To Which Department Should SDRs Report:  Sales or Marketing?

Historically, SDRs reported to sales.  That’s probably because sales first decided to fund SDR teams as a way getting inbound lead management out of the hands of salespeople [5].  Doing so would:

  • Enable the company to consistently respond in a timely manner to all inquiries
  • Free up sales to spend more time on selling
  • Avoid the problem of individual reps not processing new leads once they are “full up” on opportunities [6]

The problem is that most enterprise software sales VPs are not particularly process-oriented [7], because they grew up in a pre-industrialized era of sales [8].  In fact, nothing drives me crazier than an old-school, artisanal, deal-person CRO insisting on owning the SDR organization despite the total inability to manage it.  They rationalize:  “Oh, I can hire someone process-oriented to manage it.”  And I think:  “but what can that person learn from you [9] about how to manage it?”  And the answer is nothing.  Your desire to own it is either pure ego or simply a ploy to enrich your resume.

I’ll say again because it drives me crazy:  do not be the VP of Sales who insists on owning the SDR organization in the annual planning meeting but then shows zero interest in it for the rest of the year.  You’re not helping anyone!

As mentioned in a footnote in a prior post, I greatly prefer SDRs reporting to marketing versus sales.  Why?

  • Marketing leadgen and nurture people are metrics- and process-oriented animals, naturally suited to manage a process-oriented department.
  • It provides a simple, clear conceptual model:  marketing is the opportunity creation factory and sales is the opportunity closing machine.

In short, marketing’s job is to make opportunities.  Sales’ job is to close them.

# # #

Notes

[1] BANT = budget, authority, need, time-frame.

[2] Most early- and mid-stage startups put SDRs in their regular sales training sessions which I think does them a disservice.  Normal sales training is about selling products/solutions.  SDRs “sell” meetings.  They should not attempt to build business value or differentiation. Training them to do so tempts them to do – even when it is not their job.

[3] A typical QCR:SDR ratio is 3-4:1, though I’ve seen as low as 1:1 and as high as 6:1

[4] Believe it or not, this sometimes happens (typically when your reps are already carrying a lot of oppties).  Few things reflect worse on the company than a last-minute rescheduling of the meet-your-salesperson call. You don’t get a second chance to make a firm impression.

[5] Although most early models had wide bypass rules  – e.g.,  “leads with VP title at this list of key accounts will get passed directly to reps for qualification” – reflecting a lack of trust in marketing beyond dropping leaflets from airplanes.

[6] That problem could still exist at hand-off (i.e., opportunity creation) time but at least we have combed through the leads to find the good ones, and reports can easily identify overloaded reps.

[7] While they may be process-oriented when it comes to the sales process for a deal moving across stages during a quarter, that is not quite the same thing as a velocity mentality driven by daily or weekly goals with tracking metrics.  If you will, there’s process-oriented and Process-Oriented.

[8] One simple test:  if your sales org doesn’t have monthly cadence (e.g., goals, forecasts) then your sales VP is probably not capital P process-oriented.

[9] On the theory you should always build organizations where people can learn from their managers.