Last week, I had three separate conversations with VPs of Product about business cases… initially framed as “do you have a template we should use so that our team can prioritize big investments?” Unpacking this, it became clear that their product teams wanted to jump straight into generating numbers and spreadsheets to present to executives, but were unclear about the highest-level objectives of their various investments. In fact, their project stacks included radically different items that were hard to compare one-to-one. The hope was that running business cases (revenue numbers) would give us a uniform way of sorting cats, dogs, sheep, goldfish and elephants.
I love templates and canvases. (Anyone not familiar with Alex Osterwalder’s Business Model & Value Proposition Canvas should start there.) But tools don’t do our thinking for us. All three of my coaching conversations shifted toward clarity of objectives and audience. Said another way, business cases are stories about money and we need to know what kind of story to tell before we can crunch the right numbers.
(Hollywood genres help us find a subset of movies that we might watch: romcoms, SciFi shoot-em-ups, political documentaries, psychological thrillers, police procedurals. It’s hard to force-rank “A Night At The Opera” versus “Pan’s Labyrinth” versus “Independence Day” along a single axis. Likewise, we should compare new investments against projects with similar goals – so that we can find the “best” of each kind within a broader portfolio view. No software product is made up of only one thing.)
So before we start crunching financials, we should know which kind of money story we are telling. Who is it for? How much is it worth to them? How sure are we that they will use/buy it? Can we ballpark benefits versus costs? Some examples:
 Stories About Internal Cost Savings
Internal cost savings requests are usually about a tool or app or process that will help some of our employees do more / faster / better work. “A smarter search tool will help our Support team find solutions to frequent customer issues faster and reduce time to close tickets.” “We’re manually checking all mortgage applications that have poor credit scores because our automated approval system gets these wrong. Need to fix that!” “We’d save millions in scrap metal if our manufacturing robots had more accurate vision systems.”
A simple cost savings business case mostly writes itself:
• Annual Support hours * average Support salary * estimated % improvement = $Savings • E.g. 25,000 hours/year spent on support tickets * $70/hour burdened costs * 10% improvement = $175k guesstimated savings • If a $20k/year search tool delivers this savings, that’s a 2 month payback!
And we can reasonably compare other cost savings projects or proposals based on risk, departmental buy-in, etc. (We want Support to check our assumptions, and get their agreement to slow down hiring if this really works.)
 Stories About Upselling Current Paying Customers
Upselling means extracting more money from our existing installed base, presumably by delivering new capabilities that they value — and will pay for. “Restricting the free version of our MMORPG to 10 players and 2 hours will encourage serious gamers to pay for the unlimited version.” “Adding hybrid cloud capabilities to our Enterprise Edition will encourage Standard Edition subscribers to upgrade.” “Most of our ERP customers also need a staff scheduling application. Let’s white-label Partner X’s schedule app and resell the crap out of it!”
Notice that these money stories are about current customers, not our employees. So they raise questions about what fraction of our base would be willing to buy, and how we’ve confirmed that.
An upsell money story looks like:
• Total current customers * estimated % upgrade * upsell price • E.g. 19,000 paid subscribers * 5% upsell in first year * $3000 incremental license fees = $2.9M • If this costs us $200k/year, it’s a huge win. If we’ll spend $4M, then let’s pass.
The biggest risks are in over-estimating customer interest or willingness to pay more. If the 5% above turns out to be 0.5%, we’ve made a poor choice. So it’s essential that we get out of our own heads – or spreadsheets – for non-selling interviews with real users.
I rarely see companies compare actual results to business case predictions, however – which encourages overly aggressive business cases. And a false sense of exactness: our 5% penetration may be wildly speculative, so intellectually more honest to use a range for our riskiest variable (“2%-5%”) and throw away second digits (use $3M instead of $2.85M, and be suspicious of even that first digit).
We probably have a limit of two major installed base upsell campaigns per year. Comparing similar stories against each other lets up pick two that have good evidence plus good upside.
 Stories About Opening Up a New Market or Segment
Since these address customers we don’t already have — in markets we know less about – they tend to be more speculative. “Adding Azure support to our AWS-only cloud analytics app lets us reach an unserved audience.” “Doctors love our patient check-in solution. What about veterinarians?” “If we retarget our SMB accounting software to mid-sized banks, we can boost our deal size 20x”.
The math is simple, with much larger error bars:
• Total veterinarian practices * guesstimated penetration in year 1 * $ASP • E.g. 113k private practice vets * 3% penetration * $900/year = $3M • With error bars, 50k-100k veterinarians * 1%-3% penetration * $900 = $0.5M-$3M
I’m quite skeptical about such stories: they can be revenue-driven fever dreams based on one outlier opportunity, extrapolating from little evidence. We imagine it’s easier to sell into some unknown segment than the one we’re struggling in, and easy to skate past real product differences or market challenges.
So I ask lots of hard questions: how many prospects have we interviewed? Do we know the names of our first 15 customers? What’s different about product fit in this segment? Would we need new distribution channels, marketing campaigns, packaging/pricing, support processes, integration partners, competitive intelligence… Gearing up for a new segment is risky and expensive. We’d rather not stumble into $20M of incremental costs and a deeply defocused organization for an optimistic $1-3M.
(Long ago, we called this the China Fallacy: “There are 1B consumers in China. If we can just get 1% to buy our widget…”)
 Stories About Improved Customer Satisfaction or NPS
It’s easy to tell a qualitative story about happier users, but harder to connect that with money. That encourages sloppy investment thinking, where we chase better survey scores but never tie them back to concrete customer decisions or business results.
“We got lots of support calls for help with exporting data. Time to redesign that feature.” “Onboarding new customers is taking 65 days on average. How do we chop that to 45 or 50 days?” “NPS is 28 and needs to be 41 by Q4.” “I can’t believe how many bugs we have!”
Ideally, we can make a causal connection between Customer Sat and renewal rates (churn) or upsell or enthusiastic referrals. But these stories tend toward staying competitive, removing obvious warts, or reducing product frustration. We need to do a lot of things to keep products working, users renewing, and our company in business. So our money stories are often loss avoidance or negative revenue recapture: “10% of churned customers identified our billing problems as one reason they left.” Fixing the billing cycle or improving user onboarding are vaguely connected to increased subscription renewal rates, but less directly.
Yet we have to invest part of our effort into keeping the lights on. I typically allocate about 1/3 of total engineering effort against performance, security, design improvements and tech debt – even though it’s problematic to tie any one improvement to clear financial outcomes. (See these two posts.) Because it’s too late when we get hacked or customers complain about snail-slow response times – we have to stay ahead customer perception. Every sprint should include some bug fixes and refactoring and UX improvement. We boost quality and ease-of-use now, knowing that these eventually lead to happier users and higher renewal rates. Investing only in the current quarter is a going-out-of-business strategy. (Sales-dominant companies defer software hygiene until they close their doors.)
Now That We’ve Organized These Stories…
Of course, sorting into piles doesn’t avoid the need for hard decisions. But I find that comparing like stories helps us quickly identify the subset worthy of real business cases. Noticing that we have twenty ideas for upselling current users makes it clear that we must pick three or four for deeper analysis. Comparing a dozen cost-saving suggestions lets us eliminate the marginal ideas. Stack-ranking new market opportunities against each other highlights their relative risks and potential so that we can pick one or two for serious market validation.
Our brains can’t manage backlogs of 20 or 500 items. So rapidly identifying a few candidates lets us focus our brains where it matters. Now we’re ready to pull out canvasses and templates.
Comparing unlike items is very difficult. So applying one universal business model template to dissimilar kinds of money stories is problematic. Grouping similar money stories by audience and top-level objective is a good first step toward prioritizing requests.