Unit economics and sales and delivery projections

Last month I described the general issues underlying projections. Now it's time to drill down a bit. As a first step, a projection for selling something typically involves generating leads, converting those leads into sales, getting paid, and delivering on what you sold. None of these things are free, and none of them are instantaneous. The unit economics has to do with what everything costs on a per unit basis.

Congratulations, you have the first step done...

What if?

The most common thing we see is that the analysis stops here. But beware. What happens if your numbers are off by just a little? As it turns out, little differences can have big effects, especially when amplified by large numbers and taken through sequential processes.

Pricing is always a challenge, because you can get a higher percentage of sales at a lower price - usually. Of course higher prices make things seem more valuable and you might actually make more sales at a higher price. The only real way to find out is to test. Of course testing takes time and money too, and when you have little resources, you might have to limit yourself to quick surveys and similar methods. In terms of modeling, it's pretty easy to model at different price points and different percentages or progress through the sales sieve.

Cost is often controllable as well. Of course quality varies with cost, and if you are selling yourself as high-end, you need to have high quality or you will ultimately lose the client base to competitors with higher quality - maybe... As it turns out, quality may not be readily discernible by your customers. Selling expertise, for example, it takes an expert to know an expert, so if the customer is not an expert, they will use surrogates like institutional reputation, attire, word usage, accents, and so forth. See how certain all this stuff is?

One of the easiest things to vary to find out how sensitive your model is to changes will be the percentage going from step to step in the sales sieve. As it turns out, a small change in the percentage going from step to step make so big difference over the sales cycle and fulfillment process. It also turns out that the likelihood of moving from step to step is one of the hardest things to get right in projections, until you execute on it enough to narrow the percentages down to relatively little variance. Of course no variance is just not realistic.

I will start with a simple example

StepTime (days)Tot Days Min < Quantity < Max$/eaMin < Cost Here < MaxMin < Net To Here < MaxSuccess %+/-Min < Left < MaxNote
Obtain leads11500050005000-$1.00($5,000.00)($5,000.00)($5,000.00)($5,000.00)($5,000.00)($5,000.00)2.00%1.00%50.00100.00150.00Use database of emails, existing audiences of startups CEOs and investors in FB ad campaign
Retargeting and follow up drip campaign to Free Trial Signup151650.00100.00150.00-$4.00($200.00)($400.00)($600.00)($5,200.00)($5,400.00)($5,600.00)50.00%10.00%20.0050.0090.00Drip content to these people with each additional batch of interested prospects compounding previous month of work. Lead signs up for site (provides CC information)
Lead Converts to Paid Plan143020.0050.0090.00$100.00$2,000.00$5,000.00$9,000.00($3,200.00)($400.00)$3,400.0050.00%10.00%8.0025.0054.00Range of 15-50% (50% is for opt out trials)
Recurring30608.0025.0054.00$100.00$800.00$2,500.00$5,400.00($2,400.00)$2,100.00$8,800.0080.00%10.00%5.6020.0048.60Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Recurring30905.6020.0048.60$100.00$560.00$2,000.00$4,860.00($1,840.00)$4,100.00$13,660.0080.00%10.00%3.9216.0043.74Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Recurring301203.9216.0043.74$100.00$392.00$1,600.00$4,374.00($1,448.00)$5,700.00$18,034.0080.00%10.00%2.7412.8039.37Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Recurring301502.7412.8039.37$100.00$274.00$1,280.00$3,936.100($1,174.00)$6,980.00$21,971.0080.00%10.00%1.9210.2435.43Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Recurring301801.9210.2435.43$100.00$192.00$1,024.00$3,543.00($982.00)$8,004.00$25,514.0080.00%10.00%1.348.1931.89Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Recurring302101.348.1931.89$100.00$134.00$819.00$3,189.00($848.00)$8,823.00$28,703.0080.00%10.00%0.946.5528.70Recurring revenue $100/mo/customer. Expected dropout rate is 10-30%
Sales Sieve

Yes, this is a simple example, of a realistic projection. It might help to describe what's in it.

The columns in yellow and green indicate the results of the lower and higher percentage of customers moving to the next step. As you can see, a 10% difference in movement from step to step goes from never making money to making more than $28,000 in 210 days for each $5400 invested in the sales process. This unit economics model can be repeated with more cash investment, but as you can also see, in the middle case, after 60 days, more cash has arrived than gone out. By 120 days, there is enough money to pay for the cycle to be repeated. And that means you can double your investment in sales every 120 days, or 3 times per year. If things go well, and you end up on the high side of movement through the sieve, you can double your investment in sales after 60 days, or 6 times per year. It gets even better if you look at the ongoing process, because you keep generating additional revenue from prior sales. The recurring revenue gives even higher growth potential. But just to be clear, if you are off by 10% in the other direction, you lose money, and the more you put in, the more you lose.

We call this the envelope of anticipated futures. You start with an envelope that has a lot of variances and seek to constrain the future to a desired one. As you adapt and learn, you tighten the envelope and try to assure better future outcomes while detecting changes that make a real difference. The goal is to anticipate potential problems by projecting forward in time and mitigate them before they become kinetic problems.


I feel your pain. Now it's time to feel mine... As an investor, I don't just want to see your rosy projections of a future when everything goes right. Most projections take a single estimate and extrapolate toward the future, not taking into account the variations on what could happen. In the example above, there are three projections run simultaneously. The one in yellow is what happens when the percentage going from stage to stage is lower then the expected, and the one in green is what happens if things go better than expected. I'm going to want to know how you came to these percentages, and how close they are. And you should want to know as well.

A call to action

Building unit economics and financial projections far the sales process is something we help almost every company we advise do at some point. Advisory boards and go to market strategies always involve this sort of analysis, and they should. How do we do this? It starts with filling in some initial forms, followed by substantial discussion and updates, and ultimately leads to projections like the one above... Of course we help make sure you don't get the red parts in your projections by helping you adapt your approach for success. What's the first step?

Go To Angel

Once we start working with you, we'll start to look at your unit economics and help you discover improvements that can make all the difference.

In summary

Don't just guess - project an envelope taking into account uncertainty.

Copyright(c) Fred Cohen, 2021 - All Rights Reserved