I’ve mentioned this idea a few times of late (e.g., my previous post, my SaaStock EMEA presentation)  and I’ve had some follow-up questions from readers, so I thought I’d do a quick post on the subject.
Back in the day at Salesforce, we called pipeline sources “horsemen,” a flawed term both for its embedded gender pronoun and its apocalyptic connotation. Nevertheless, for me it did serve one purpose — I always remembered there were four of them.
Today, I call them “pipeline sources” but I’ve also heard them referred to as “pipegen sources” (as in pipeline generation) and even “revenue engines” which I think is an over-reach, if not a well intentioned one .
While you can define them in different ways, I think a pretty standard way of defining the pipeline sources is as follows:
- Marketing, also known as “marketing/inbound.” Opportunities generated as a result of people responding to marketing campaigns .
- SDRs, also known as “SDR/outbound,” to differentiate these truly SDR-generated oppties from marketing/inbound oppties that are also processed by SDRs, but not generated by them .
- Alliances . Opportunities referred to the company by partners, for example, when a regional system integrator brings the company into a deal as a solution for one of its customers.
- Sales, also known as “sales/outbound,” when a quota-carrying salesrep does their own prospecting, typically found in named-account territory models, and develops an opportunity themselves.
Product-led growth (PLG) companies should probably have a fifth source, product, but I won’t drill into PLG in this post [5A].
Attribution issues (i.e., who gets credit when an opportunity is developed through multiple touches with multiple contacts over multiple quarters  ) are undoubtedly complex. See note  not for the answer to the attribution riddle, but for my advice on best dealing with the fact that it’s unanswerable.
Now, for the money question: what’s the right allocation across sources? I think the following are reasonable targets for a circa $50M enterprise SaaS company for mix of oppties generated by each source (all targets are plus-or-minus 10%):
- Marketing: 60%
- SDR/outbound: 10%
- Alliances: 20%
- Sales/outbound: 10%
Now, let’s be clear. This can vary widely. I’ve seen companies where marketing generates 95% of the pipeline and those where it generates almost none. SDR/outbound makes the most sense in a named-account sales model, so I personally wouldn’t recommend doing outbound for outbound’s sake  . Alliances is often under 20%, because the CEO doesn’t give them a concrete oppty-generation goal (or because they’re focused more on managing technology alliances). Sales/outbound only makes sense for sellers with named-account territories, despite old-school sales managers’ tendency to want everyone prospecting as a character-building exercise.
And let’s not get so focused on the mix that we forget about the point: cost-effective opportunity generation (ultimately revealed in the CAC ratio) with broad reach into the target market.
Now, for a few pro tips:
- Assign the goal as a number of oppties, not a percentage. For example, if you want 60% from marketing and have an overall goal of 100 oppties, do not set marketing’s goal at 60%, tell them you want 60 oppties. Why? Because if the company only generates 50 oppties during the quarter and marketing generates 35 of those, then marketing is popping champagne for generating 70% of the oppties (beating the 60% goal), while they are 15 oppties short of what the company actually needed.
- Use overallocation when spinning up new pipeline sources. Say you’ve just created an RSI alliances team and want them generating 10% of oppties. By default, you’ll drop marketing’s target from 70% to 60% and marketing will build a budget to generate 60% (of say 100) oppties, so 60 oppties. If they need $3K worth of marketing to generate an oppty, then they’ll ask for $180K of demandgen budget. But what if alliances flames out? Far better to tell marketing to generate 70 oppties, give them $210K in budget to do so and effectively over-assign oppty generation to an overall goal of 110 when you need 100. This way, you’re covered when the new and presumably unpredictable pipeline generation source is coming online .
# # #
 Video forthcoming if I can get access to it.
 The good intentions are to keep everyone focused on revenue. The over-reach is they’re not really engines, more fuel sources. I am a big believer in the concept of “revenue engines,” but I use the term to refer to independent business units that have an incremental revenue target and succeed or fail in either an uncoupled or loosely coupled manner. For example, I’d say that geographic units (e.g., Americas, EMEA), channels (e.g., OEM, VAR, enterprise sales, corporate sales), or even product lines (depending on the org) are revenue engines. The point of having revenue engines is diversification, as with airplanes, they can sputter (or flame-out) independently. (As one aviation pioneer was reputed to have said: “why do I only fly four-engine planes across the Atlantic? Because they don’t make five-engine planes.”)
 I will resist the temptation to deep dive into the rabbit hole of attribution and say two things: (a) you likely have an attribution mechanism in place today and (b) that system is invariably imperfect so you should make sure you understand how it works and understand its limitations to avoid making myopic decisions. For example, if an oppty is created after several people downloaded a white paper, a few attended a webinar, an SDR had been doing outreach in the account, the salesperson met a contact on the train, and a partner was trying to win business in the account, who gets the credit? It’s not obvious how to do this correctly and if your system is “one oppty, one source” (as I’d usually recommend over some point allocation system), there will invariably be internal jockeying for the credit.
 SDRs are often split inbound vs. outbound not only to ease the tracking but because the nature of the work is fundamentally different. Hybrid SDR roles are difficult for this reason, particularly in inbound-heavy environments where there is always more inbound work to do.
 My taxonomy is that there are two types of “partners” — “channels” who sell our software and “alliances” who do not. In this case (where we’re talking about pipeline generation for our direct salesforce), I am speaking of alliance partners, who typically work in a co-sell relationship and bring the company into oppties as a result. In the case of channels, the question is one of visibility: are the channels giving us visibility into their oppties (e.g., in our CRM) as you might find with RSIs or are they simply forecasting a number and mailing us a royalty check as you might find with OEMs.
[5A] Product meaning trials (or downloads in open source land), which effectively become the majority top-of-funnel lead source for PLG companies. This begs the question: who drives people to do those trials (typically marketing and/or word of mouth)
 One simple, common example: a person downloads a white paper they found via through a search advertisement five quarters ago, ends up in our database, receives our periodic newsletter, and then is developed by an SDR through an outreach sequence. Who gets the credit for the opportunity? Marketing (for finding them in the first place and providing a baseline nurture program via the newsletter) or SDR/outbound (for developing them into an oppty)? Most folks would say SDR in this case, but if your company practices “management by reductio ad absurdum” then someone might want to shut down search advertising because it’s “not producing” whereas the SDRs are. Add some corporate politics where perhaps sales is trying to win points for showing how great they are at managing SDRs after having taken them from marketing and things can get … pretty icky.
 Another favorite example: marketing sponsors a booth at the Snowflake user conference and we find a lead that develops into an opportunity. Does marketing get the credit (because it’s a marketing program) or alliances (because Snowflake’s a partner). Add some politics where the alliances team has been seen as underperforming and really needs the credit, and things can get again yucky and confusing, leading you away from the semi-obvious right answer: marketing, because they ran a tradeshow booth and got a lead. If you don’t credit marketing here, you are disincenting them from spending money at partner conferences (all I, no RO.) The full answer here is, IMHO, to credit marketing with being the source of oppty, to track influence ARR by partner so we know how much of our business happens with which partners, and to not incent the technology alliances group with opportunity creation targets. (Oppty creation, however, should be an important goal for the regional and/or global system integrator alliances teams.)
 My recommended solution here is two-fold: (a) use whatever attribution mechanism you want, ensuring you understand its limitations, and (b) perform a win-touch analysis at every QBR where a reasonably neutral party like salesops presents the full touch history for a set of representative deals (and/or large) deals won in the prior quarter. This pulls everyone’s heads of our their spreadsheets and back into reality — and should ease political tensions as well.
 Having an SDR convince someone to take a meeting usually results in a higher no-show rate and a lower overall conversion rate than setting up meetings with people who have engaged with our marketing or our partners already.
 Put differently, you should stalk customers only when you’re quite sure they should buy from you, but they haven’t figured that out yet.
 And yes there’s no free lunch here. Your CAC will increase because you’re paying to generate 110 oppties when you only need 100. But far better to have the CAC kick up a bit when you’re starting a new program than to miss the number because the pipeline was insufficient.
Dave, I have always followed your advice to identify and measure the 4 primary sources of pipeline. As you mentioned, percentage of pipeline by source is highly variable based upon multiple variables, including Go-To-Market motion (sales led versus sales led, inside sales versus field sales, solution ACV, etc.).
Our research has captured the % pipeline generated by source (Sales Development, Direct Sales (AEs), and inbound marketing leads in recent research. We also have captured it at the Account-Based Marketing program level. The trends over the last few years has dramatically shifted pipeline contribution from marketing to sales development – but that is highly variable versus how “attribution” is factored into the percentages ;-(
Hit me with some numbers Ray! And yes, first question we’d both have was did the sales or SDRs truly generate the opportunity or were they calling into lists created and nurtured by marketing programs? Key point to me is not to get “over excited” about crediting people for this stuff. See my posts on “what really matters in sales” and it’s too things: pipeline coverage and pipeline conversion. We need to make sure we give ourselves a chance to hit the number and pipegen is all about that. So yes attribution is key and given its limitations we can’t get to excited about giving big bonuses or promotions based on it — we might be rewarding gamespersponship over performance. Either way, I’d love a few numbers!
How would you attribute customer referrals?
Thanks for asking! At the risk of going into the attribution rabbit hole … well, for me word of mouth (WoM) is the lead source and if I had lead that converted to an oppty and the lead source was WoM then I’d attribute it to the marketing funnel. The deeper issue this cuts to is the limits of “our model” vs. “their reality” so I would personally try, where possible, to ask “how did you hear about us?” even if I theoretically knew their lead source (e.g., search) … because I want to know what made them run the search.
Dave, love this, especially the money answer. How do you see this change as a company grows/matures?
Thanks for reading and I hope you’re well, Josh. That could be a whole post unto itself. Rough / quick guestimate answer: early (80, 0, 0, 10), growth (60, 10, 20, 10), scale (50, 10, 30, 10) — but there a whole lot of underlying assumptions in there about business model, GTM strategy, ABM. Net is marketing goes down, alliances goes up and the other two say roughly constant.
Customer Success can also be a source of pipeline. Of course the attribution needs to work with AEs and Success not fighting for credit.
yes and yes