Years ago, I was in a board meeting and a VC dropped the expression: “you can’t eat IRR.”
I’d never heard it before. It sounded catchy. But, honestly, I didn’t know what it meant. At first, I thought it was VC-ism. But then I learned the phrase was coined by the venerable Howard Marks of Oaktree in one of his famous investor memos. So it’s really a finance-ism, not a VC-ism.
What does it mean? While Marks’ fourteen-page memo provides numerous clear, in-depth examples, I’ll try to capture the spirit of the question with one of my own. Consider two investments, A and B:
And let’s pretend they’re both made by a VC firm on your behalf. Firm A puts your $100 into a startup and one year later they sell the company and distribute $120 back to you. The internal rate of return (IRR) is 20%. That’s pretty good (e.g., compared to average stock market returns), but you were hoping for a long-term investment and now you need to find somewhere else to put your money. Firm B invests in a startup and sells it in year eight, getting you $350 back. The IRR is also 20%. But the total value to paid in (TVPI) is 3.5 compared to 1.2 with Firm A.
In essence, because IRR is based on time, if you can produce a nice return very quickly, you can get an amazing IRR. In my example, both investments produced a solid IRR of 20%, but in one case I had only $20 of profit to eat with, whereas with the other I had $250. (And this is why most investors prefer TVPI over IRR as their top metric. Why? Because you can’t eat IRR.)
I feel the same way about pipegen (aka, pipeline generation). Why? Let’s look at a simple example:
This company’s pipegen targets, presumably created using some funnel model, are in the first row. That model is typically some kind of waterfall where you take starting pipeline, add pipegen plans from the various pipegen sources (e.g., marketing, AEs, SDR, alliances), subtract closed/won opportunities, subtract lost and slipped opportunities, and make some adjustments for opportunity sizes varying around over time.
Actual pipegen, expressed as a percent of plan, is in second row. This is usually broken out by source, but I’ve aggregated them here to keep things simple. Note that if you stopped reading after row two, you’d pop the champagne. Go us! 105 to 109% of pipegen targets all year!! We’re the best!
And you’d be surprised how many companies do this. Sometimes they do it unknowingly. Once you break out pipegen by source, look at plan performance, and compare to prior quarters and years, you have a lot of numbers on a slide. So you tend to then look at the bottom and see total pipgen as a percent of target. If those figures are 100%+, then awesome. Pipegen’s not the problem. Next slide.
But that’s not good enough. What if, for whatever reason, that despite our strong pipegen performance relative to plan, that we’re not starting each quarter with sufficient pipeline coverage. Remember, there are only two high-level questions about sales:
- Did we give them the chance to make the number?
- Did they make the number?
Look at row 3. Starting coverage is short for starting pipeline coverage. That’s total current-quarter pipeline at the start of the quarter divided by the new ARR target for the quarter. Most CROs want this ratio to be 3.0x. Here, you can see we started every quarter with between 2.5x to 2.8x coverage. That’s not good enough. In short, it means that we’re not giving sales a fair shot at making the number each quarter. That might partially be sales’ fault. Normally sales is responsible for generating 20-40% of the pipeline (including SDR outbound). But it’s also the fault of marketing and alliances. Nobody should be drinking champagne.
Sure, we may have beaten the pipegen targets in our model, but something’s clearly wrong with our model if we can consistently miss starting coverage while exceeding pipegen targets. What might that be? Perhaps:
- We’re losing more deals so less pipeline is slipping
- Deals are shrinking, e.g., in response to price pressure from a competitor
- Sales cycles are stalled by a megavendor entering the space
- Deals are slipping more frequently and/or by more quarters than in the past
- Sales cycles are lengthening, so we’re generating enough pipeline but it’s all too far in the future
- There’s a math mistake in our model
There are a lot of different reasons this could happen. But the main point is don’t forget to look at row three in the example. That’s where the pipegen team can celebrate. They should need two triggers to open the champagne: first that we beat pipegen targets and second that we start with sufficient pipeline coverage.
And we only break out the caviar on the third trigger: when sales beats the new ARR number.
This post is similar in spirit to one I did last fall entitled, Why Great Marketers Look at Pipeline Coverage, Not Just Pipeline Generation. If you want more on this topic, then take a look at that post as well.
Finally, you now know why I say, “you can’t eat pipegen.” But you can eat starting coverage.
