The Top Two, High-Level Questions About Sales (and Associated Metrics)

“The nice thing about metrics is that there are so many to choose from.” — Adapted from Grace Hopper [1]

“Data, data everywhere.  Nor any drop to drink.” — adapted from Samuel Taylor Coleridge [2]

In a world where many executives are overwhelmed with sales and marketing metrics — from MQL generation to pipeline analysis to close-rates and everything in between — I am writing this post in the spirit of kicking it back up to the CXO-level and answering the question:  when it comes to sales, what do you really need to worry about?

I think can burn it all down to two questions:

• Are we giving ourselves the chance to hit the number?
• Are we hitting the number?

That’s it.  In slightly longer form:

• Are we generating enough pipeline so that we start every quarter with a realistic chance to make the number?
• Are we converting enough of that pipeline so that we do, in fact, hit the number?

Translating it to metrics:

• Do we start every quarter with sufficient pipeline coverage?
• Do we have sufficient pipeline conversion to hit the number?

Who Owns Pipeline Coverage and How to Measure It?
Pipeline coverage is a pretty simple concept:  it’s the dollar value of the pipeline with a close date in a given period divided by the new ARR target for that period.  I have written a lot of pretty in-depth material on managing the pipeline in this blog and I won’t rehash all that here.

The key points are:

• There are typically four major pipeline generation (pipegen) sources [3] and I like setting quarterly pipegen goals for each, and doing so in terms of opportunity (oppty) count, not pipeline dollars.  Why?  Because it’s more tangible [4] and for early-stage oppties one is simply a proxy for the other — and a gameable one at that [5].
• I loathe looking at rolling-four-quarter pipeline both because we don’t have rolling-four-quarter sales targets and because doing so often results in a pipeline that resembles a Tantalean punishment where all the deals are two quarters out.
• Unless delegated, ownership for overall pipeline coverage boomerangs back on the CEO [6].  I think the CMO should be designated the quarterback of the pipeline and be responsible for both (a) hitting the quarterly goal for marketing-generated oppties and (b) forecasting day-one, next-quarter pipeline and taking appropriate remedial action — working across all four sources — to ensure it is adequate.
• A reasonable pipeline coverage ratio is 3.0x, though you should likely use your historical conversion rates once you have them. [7]
• Having sufficient aggregate pipeline can mask a feast-or-famine situation with individual sellers, so always keep an eye on the opportunity histogram as well.  Having enough total oppties won’t help you hit the sales target if all the oppties are sitting with three sellers who can’t call everyone all back.
• Finally, don’t forget the not-so-subtle difference between day-one and week-three pipeline [8].  I like coverage goals focused on day-one pipeline coverage [9], but I prefer doing analytics (e.g., pipeline conversion rates) off week-three snapshots [10].

Who Owns Pipeline Conversion and How to Measure and Improve It?
Unlike pipeline coverage, which usually a joint production of four different teams, pipeline conversion is typically the exclusive the domain of sales [11].  In other words, who owns pipeline conversion?  Sales.

My favorite way to measure pipeline conversion is take a snapshot of the current-quarter pipeline in week 3 of each quarter and then divide the actual quarterly sales by the week 3 pipeline.  For example, if we had $10M in current-quarter new ARR pipeline at the start of week 3, and closed the quarter out with $2.7M in new ARR, then we’d have a 27% week 3 pipeline conversion rate [12].

What’s a good rate?  Generally, it’s the inverse of your desired pipeline coverage ratio.  That is, if you like a 3.0x week 3 pipeline coverage ratio, you’re saying you expect a 33% week 3 pipeline conversation rate.  If you like 4.0x, you’re saying you expect 25% [13].

Should this number be the same as your stage-2-to-close (S2TC) rate?  That is, the close rate of sales-accepted (i.e., “stage 2” in my parlance) oppties.  The answer, somewhat counter-intuitively, is no.  Why?

• The S2TC rate is count-based, not ARR-dollar-based, and can therefore differ.
• The S2TC rate is typically cohort-based, not milestone-based — i.e., it takes a cohort of S2 oppties generated in some past quarter and tracks them until they eventually close [14].

While I think the S2TC rate is a better, more accurate measure of what percent of your S2 oppties (eventually) close, it is simply not the same thing as a week-3 pipeline conversion rate [15].  The two are not unrelated, but nor are they the same.

There are a zillion different ways to improve pipeline conversion rates, but they generally fall into these buckets:

• Generate higher-quality pipeline.  This is almost tautological because my definition of higher-quality pipeline is pipeline that converts at a higher rate.  That said, higher-quality generally means “more, realer” oppties as it’s well known that sellers drop the quality bar on oppties when pipeline is thin, and thus the oppties become less real.  Increasing the percent of pipeline within the ideal customer profile (ICP) is also a good way of improving pipeline quality [16] as is using intent data to find people who are actively out shopping.  High slip and derail percentages are often indicators of low-quality pipeline.
• Make the product easier to sell.  Make a series of product changes, messaging/positioning changes, and/or create new sales tools that make it easier to sell the product, as measured by close rates or win rates.
• Make seller hiring profile improvements so that you are hiring sellers who are more likely to be successful in selling your product.  It’s stunning to me how often this simple act is overlooked.  Who you’re hiring has a huge impact on how much they sell.
• Makes sales process improvements, such as adopting a sales methodology, improving your onboarding and periodic sales training, and/or separating out pipeline scrubs from forecast calls from deal reviews [17].

Interestingly, I didn’t add “change your sales model” to the list as I mentally separate model selection from model execution, but that’s admittedly an arbitrary delineation.  My gut is:  if your pipeline conversion is weak, do the above things to improve execution efficiency of your model.  If your CAC is high, re-evaluate your sales model.  I’ll think some more about that and maybe do a subsequent post [18].

In conclusion, let’s zoom it back up and say:  if you’ve got a problem with your sales performance, there are really only two questions you need to focus on.  While we (perhaps inadvertently) demonstrated that you can drill deeply into them — those two simple questions remain:

• Are we giving ourselves the chance to hit the number?
• Are we hitting it?

The first is about pipeline generation and coverage.  The second is about pipeline conversion.

# # #

Notes

[1]  The original quip was about standards:  “the nice thing about standards is that you have so many to chose from.”

[2]  The original line from The Rime of the Ancient Mariner was about water, of course.

[3]  I remember there are four because back in the day at Salesforce they were known, oddly, as the “four horsemen” of the pipeline:  marketing, SDR/outbound, alliances, and sales.

[4]  Think:  “get 10 oppties” instead of “get $500K in pipeline.”

[5]  Think:  ” I know our ASP is $50K and our goal was $500K in pipeline, so we needed 10 deals, but we only got 9, so can you make one of them worth $100K in the pipeline so I can hit my coverage goal?”  Moreover, if you believe that oppties should be created with $0 value until a price is socialized with the customer, the only thing you can reasonably measure is oppty count, not oppty dollars.  (Unless you create an implied pipeline by valuing zero-dollar oppties at your ASP.)

[6]  Typically the four pipeline sources converge in the org chart only at the CEO.

[7]  And yes it will vary across new vs. expansion business, so 3.0x is really more of a blended rate.  Example:  a 75%/25% split between new logo and expansion ARR with coverage ratios of 3.5x and 1.5x respectively yields a perfect, blended 3.0 coverage ratio.

[8]  Because of two, typically offsetting, factors:  sales clean-up during the first few weeks of the quarter which tends to reduce pipeline and (typically marketing-led) pipeline generation during those same few weeks.

[9]  For the simple reason that we know if we hit it immediately at the end of the quarter — and for the more subtle reason that we don’t provide perverse disincentives for cleaning up the pipeline at the start of the quarter.  (Think:  “why did your people push all that stuff out the pipeline right before they snapshotted it to see if I made my coverage goal?”)

[10]  To the extent you have a massive drop-off between day 1 and week 3, it’s a problem and one likely caused by only scrubbing this-quarter pipeline during pipeline scrubs and thus turning next-quarter into an opportunity garbage dump.  Solve this problem by doing pipeline scrubs that scrub the all-quarter pipeline (i.e., oppties in the pipeline with a close date in any future quarter).  However, even when you’re doing that it seems that sales management still needs a week or two at the start of every quarter to really clean things up.  Hence my desire to do analytics based on week 3 snapshots.

[11] Even if you rely on channel partners to make some sales and have two different sales organizations as a result, channel sales is still sales — just sales using a different sales model one where, in effect, channel sales reps function more like direct sales managers.

[12]  Technically, it may not be “conversion” as some closed oppties may not be present in the week 3 pipeline (e.g., if created in week 4 or if pulled forward in week 6 from next quarter).  The shorter your sales cycle, the less well this technique works, but if you are dealing with an average sales cycle of 6-12 months, then this technique works fine.  In that case, in general, if it’s not in the pipeline in week 3 it can’t close.  Moreover, if you have a long sales cycle and nevertheless lose lots of individual oppties from your week 3 pipeline that get replaced by “newly discovered” (yet somehow reasonably mature oppties) and/or oppties that inflate greatly in size, then I think your sales management has a pipeline discipline problem, either allowing or complicit in hiding information that should be clearly shown in the pipeline.

[13]  This assumes you haven’t sold anything by week 3 which, while not atypical, does not happen in more “linear” businesses and/or where sales backlogs orders.  In these cases, you should look at to-go coverage and conversion rates.

[14]  See my writings on time-based close rates and cohort- vs. milestone-based analysis.

[15] The other big problem with the S2TC rate is that it can only be calculated on a lagging basis.  With an average sales cycle of 3 quarters, you won’t be able to accurately measure the S2TC rate of oppties generated in 1Q21 until 4Q21 or 1Q22 (or even later, if your distribution has a long tail — in which case, I’d recommend capping it at some point and talking about a “six-quarter S2TC rate” or such).

[16]  Provided of course you have a data-supported ICP where oppties at companies within the ICP actually do close at a higher rate than those outside.  In my experience, this is usually not the case, as most ICPs are more aspirational than data-driven.

[17]  Many sales managers try to run a single “weekly call” that does all three of these things and thus does each poorly.  I prefer running a forecast call that’s 100% focused on producing a forecast, a pipeline scrub that reviews every oppty in a seller’s pipeline on the key fields (e.g., close date, value, stage, forecast category), and deal reviews that are 100% focused on pulling a team together to get “many eyes” and many ideas on how to help a seller win a deal.

[18] The obvious counter-argument is that improving pipeline conversion, ceteris paribus, increases new ARR which reduces CAC.  But I’m sticking by my guns for now, somewhat arbitrarily saying there’s (a) improving efficiency on an existing sales model (which does improve the CAC), and then there’s (b) fixing a CAC that is fundamentally off because the company has the wrong sales model (e.g., a high-cost field sales team doing small deals).  One is about improving the execution of a sales model; the other is about picking the appropriate sales model.