SAP design chief to his CFO: here is how design beat engineering for the next million

Your CFO does not want another deck on why design matters. He wants to know if the next million goes to design or to engineering. That question is the whole ballgame now, and the way you answer it decides whether your headcount survives the next budget cut. Let me catch you up.

The deep cut

• Translate design into finance's own math. Arin Bhowmick won his CFO by using payback period and rate of return, not craft.
• Conversion is revenue you already paid to acquire. Baymard found 70% of carts die at checkout, with billions recoverable through design fixes.
• Trust is the metric that decides who stays. MeasuringU's auto sites scored well on usability but cratered on trust at the 34th percentile.

Speak the language finance already uses

Arin Bhowmick, SAP's chief design officer, got one line from a CFO that reset his whole approach: "I don't need another presentation explaining why design is important. I need to know whether investing another million dollars in design will outperform investing it in engineering or marketing." That is the real question, and it has a real answer.

Every request a CFO sees gets two questions. How fast does the money come back, and how much does it make beyond that. Payback period and rate of return. Bhowmick's move was to judge design by those same two numbers, and to bring proof. A Forrester study for UserTesting tracked six companies and found user research returned 415% over three years and paid for itself in under six months. Finance already has the tools to grade design fairly. You just have to hand over your numbers in their format.

Where the money is already sitting

Start with conversion, because the attribution is clean and the win shows fast. Marketing pays to the cent to bring a visitor in. All of that money is spent by the time someone loads a cart. What happens after the click usually gets treated as a cosmetic layer with no owner.

The numbers make the point. Baymard Institute found seven out of ten carts die at checkout, an average of 70.22%. Fixing known checkout issues can lift conversion by 35.26%, and they estimate $260 billion in lost orders is recoverable in the US and EU through those fixes alone. Nobody had counted checkout friction as lost revenue. Point your team there first.

Trust is the number that moves the deal

Usability is table stakes, and your team is probably already good at it. The gap is trust, and that is where buyers decide to stay or leave. MeasuringU benchmarked 11 third-party auto sites and found usability averaged at the 62nd percentile while trust sank to the 34th. The average auto site does worse on trust than two-thirds of all websites in their database.

Watch what killed it. TrueCar users got hit by "immediate bombardment of phone calls and emails" after sharing contact info. Autotrader buyers said search was "too influenced by dealers." Carvana landed at a brutal −23 NPS. Kelley Blue Book led on trust, and the reason was plain: "I can check my car's value without having to provide any personal information." When users do not trust the data, they abandon. Dark patterns that juice a short-term number are burning the metric that decides loyalty.

Reuse pays twice, and so does killing bad ideas

Two more lines that survive a budget review. A design system compounds like capital. You build a component once and it pays back every time someone reuses it, and finance can verify it because the hours sit in timesheets. Sparkbox ran a controlled test where eight developers built the same page twice: 4.2 hours from scratch, 2 hours with a design system. Scale that across a year of shipping and it adds up to weeks.

Then there is research, which Bhowmick pitches as "capital protection." A prototype in front of users costs one designer's time and can kill a bad idea in weeks. Skip it and you find out at launch, after months of build. The savings never show up in a report because nobody logs the disaster that did not happen. That is exactly why research looks easy to cut and exactly why cutting it costs more later.

Three questions for your team

• Which design metric can we report on the same schedule as finance reports theirs, starting this quarter?
• Where are we trading a short-term conversion bump for lost trust, the way TrueCar and Carvana did?
• What is one bad idea we can test and kill with a prototype this month, before it eats a build cycle?