CAC Payback: 8 Months or 20? Build the 2027 Budget

Derek SimmonsSame question, same size of company, under five million in ARR. How long to pay back what you spent winning a customer? One report says about eight months.

Elena ReyesAnd the other says twenty.

Derek SimmonsTwenty. Somebody's budget is wrong, and it might be yours.

Elena ReyesOh man, that's not a rounding error. That's the difference between hiring in January and hiring in, what, next September?

Derek SimmonsRoughly, yeah. I'm Derek, this is Elena, and this is ARR Autopsy, where this week we skip the guest and build a spreadsheet.

Elena ReyesNo guest, just us, a blank tab, and a lot of opinions about cell formatting.

Derek SimmonsAnd we'll tell you which figures are sourced and which ones we invented just to make the spreadsheet run.

Elena ReyesInvented ones get a warning label. Sourced ones get a caveat if they've earned it.

Derek SimmonsGrowth does the same thing. Median growth for companies between one and five million in ARR gets quoted anywhere from forty percent to seventy-four percent.

Elena ReyesSeventy-four?

Derek SimmonsSeventy-four. And payback runs from eight and twelve months up to that twenty.

Elena ReyesSo here's my bet. The spread isn't noise. The spread is the finding.

Derek SimmonsFine, we'll treat it that way. And the end product is a twenty-twenty-seven budget you can defend to a board, which means we have to start somewhere. First, the low end. Where does the eight-month number come from?

Elena ReyesYes, please. I wanna see who's that optimistic.

Derek SimmonsOkay, so get this. HighAlpha's twenty-twenty-five SaaS Benchmarks report gives median CAC payback by ARR band, and the bands aren't labeled the way you'd hope.

Elena ReyesMeaning?

Derek SimmonsThe report lists five medians with ranges, five months with a range of two to eight, then eight, range five to fourteen, then fourteen, then twenty, then seventeen. Secondary write-ups map those five to under a million, one to five million, five to twenty, twenty to fifty, and fifty plus.

Elena ReyesSo that twenty is the twenty to fifty million band, companies far bigger than anything we're budgeting for.

Derek SimmonsRight. It's just the fourth band of the same report.

Elena ReyesSo the band we care about is the second one.

Derek SimmonsRight. For one to five million in ARR, the median is about eight months.

Elena ReyesNotice the phrase secondary write-ups. The mapping isn't printed on the chart we're staring at.

Derek SimmonsFair. That's a caveat, and I'm keeping it. The medians come from HighAlpha's own report. The labels come from people reading it.

Elena ReyesOkay, and what is that number measuring? Because payback is one of those words that means six things.

Derek SimmonsThe report defines it as months of subscription gross margin needed to recover fully loaded sales and marketing cost.

Elena ReyesGross margin, not revenue.

Derek SimmonsGross margin, so a dollar of new subscription revenue doesn't pay you back a dollar.

Elena ReyesOh, that's a nastier clock than the revenue version.

Derek SimmonsYep, and fully loaded means the whole S&M bill, not just ad spend.

Elena ReyesWhich is where founders cheat. Leave out the salaries, and you get a gorgeous number.

Derek SimmonsA gorgeous imaginary number.

Elena ReyesHa. Now, the part I care about. Eight is a median. The range for that band runs five to fourteen.

Derek SimmonsYeah.

Elena ReyesSo if I'm a company sitting at the edge of the band, say near fourteen, am I a failure or am I typical?

Derek SimmonsWhich edge? Because a company near the five million line is probably comparing itself against a very different peer set than one just past a million.

Elena ReyesExactly. Fourteen months is nearly triple the low end of the same band. So the median tells you where the middle is. It doesn't tell you whether your motion belongs in the middle.

Derek SimmonsAnd the report doesn't treat payback alone, right?

Elena ReyesNo, it pairs net revenue retention with payback as the main growth predictor.

Derek SimmonsOh, so fast payback with leaky retention is a different story from slow payback with customers who expand. Mm, I'd want both numbers on the same sheet before I believed either.

Elena ReyesNoted. Spreadsheet gets two columns.

Derek SimmonsAnd a third thing worth writing down, the range itself. Five to fourteen is the honest answer to how much this varies, and the median is just the center of it.

Elena ReyesSo the cell holds a range, not a point.

Derek SimmonsOne more thing about that eight. It's a size-based number, so it tells you about companies at your scale, not about your pricing or who you sell to.

Elena ReyesWhich is the gap the next two sources come at from the other direction. Fine, I'm ready for the higher numbers.

Derek SimmonsThen onto Benchmarkit and KeyBanc and why they sit higher. Okay, so Benchmarkit. A write-up on Marqeable summarizes the twenty-twenty-six edition, calendar year twenty-twenty-five. One hundred and ninety-eight companies in the sample. Median payback there is sixteen months.

Elena ReyesSixteen? So we've left the neighborhood of eight entirely.

Derek SimmonsWe have. And that same write-up cuts it by quartile. Top quartile, about ten months. Bottom quartile, about twenty-four.

Elena ReyesHang on. Is that top quartile directly comparable to the eight from the other report?

Derek SimmonsNot cleanly. The two sources cut their data differently, so I wouldn't line those numbers up one for one.

Elena ReyesNoted. No budget memo with that sentence in it.

Derek SimmonsAlso, look at the spread inside that one sample. Ten to twenty-four is a fourteen-month swing across one hundred and ninety-eight companies. The median is the middle, not a promise about where you land.

Elena ReyesSo a sample that size is a decent anchor, but your own number could sit anywhere across that swing.

Derek SimmonsCaveat, and I'm flagging it now, that's a secondary source, somebody else's summary of Benchmarkit. Verify it against the real report before it touches a spreadsheet.

Elena ReyesFair. What does Benchmarkit's own page say?

Derek SimmonsNot much on payback. Their public summary says twenty-twenty-five was a year efficiency improved, a blended CAC ratio of a dollar thirty, and a magic number above one.

Elena ReyesWait, a dollar thirty to acquire a dollar of new ARR? That sounds almost polite.

Derek SimmonsWildly civilized. And the page credits go-to-market rationalization, not higher spend.

Elena ReyesMeaning companies got leaner at the same job.

Derek SimmonsThat's the framing, but the payback detail is thin, so you can't lean on that page for a target.

Elena ReyesAnd the magic number above one point O, is that the good side?

Derek SimmonsThe page presents the year as efficiency improving, so I'd read it that way. But a ratio on a summary page isn't a payback target. It tells you the market got leaner, not what your sell should say.

Elena ReyesOkay, so that's a sixteen-month median we can't fully check. Now the big one.

Derek SimmonsKeyBanc and Sapphire Ventures. Their twenty-twenty-four survey, the fifteenth annual, puts the overall median CAC payback at about twenty months.

Elena ReyesAbout?

Derek SimmonsGood ear. The twenty-twenty-four number is marked as an estimate in the survey table, so treat it as a pencil figure.

Elena ReyesPencil, noted. And before?

Derek SimmonsKeyBanc's overall group was at twenty-five months in twenty-twenty-two, so it's come down five months, but it's still the slow end of everything we've heard.

Elena ReyesIs that a trend, though? Give me the middle point before I draw a line through it.

Derek SimmonsThe survey shows twenty-one in twenty-twenty-three, so twenty-five, then twenty-one, then about twenty. It's moving the right way, but the last point is the estimate, so I wouldn't call that a forecast.

Elena ReyesWhich makes me ask who is in that survey, because twenty months from a bunch of tiny companies would be alarming.

Derek SimmonsYeah. The survey's median ARR is somewhere around twenty-two million to thirty-two million dollars.

Elena ReyesWhoa, so that's not a one to three million dollar company's peer group. That's a company with a finance team and a procurement department.

Derek SimmonsRight. It skews well above the size we're budgeting for.

Elena ReyesSo one report says eight, and its band fits us. One says sixteen, and I can't check it. One says twenty, and it's measuring companies ten times our size.

Derek SimmonsSounds like three rulers and no agreement on what a foot is.

Elena ReyesStealing that. So who's right?

Derek SimmonsPossibly neither, and that's where it turns. Okay, so let's start with the cuts. HighAlpha slices by ARR bands, Benchmarkit slices by ACV bands, and KeyBanc is a much larger sample. Those cuts are one likely reason for part of the gap.

Elena ReyesPart, not all.

Derek SimmonsPart. These sources aren't directly comparable, and anyone telling you they line up is selling a slide.

Elena ReyesFine, show me the Benchmarkit slicing, because I want to see how much ACV moves it.

Derek SimmonsThe SaaS CFO site reproduces the twenty-twenty-six quartiles by ACV band. Secondary copy, so check the original. Under five thousand ACV, median payback is about eleven months.

Elena ReyesThat's close to the HighAlpha neighborhood.

Derek SimmonsRight. Five to ten thousand, twelve months. Then ten to twenty-five thousand jumps to eighteen.

Elena ReyesWhoa, okay.

Derek SimmonsTwenty-five to fifty, seventeen. Fifty to a hundred, twenty-two. Above a hundred thousand, eighteen.

Elena ReyesHold on, that's not a staircase.

Derek SimmonsNope. It goes up, dips, jumps, then drops back. I wouldn't tell a causal story from the table. The only safe read is that your ACV changes the row you land in.

Elena ReyesSo the sixteen headline is a median over a lot of different businesses.

Derek SimmonsYes. If you're selling a four thousand dollar product, that sixteen is a stranger's number.

Elena ReyesMake it concrete. Hypothetical product, eight thousand a year per customer. Which row do you use?

Derek SimmonsFive to ten thousand, so about twelve months. Now, hypothetically, raise the price to thirty thousand and you're in the twenty-five to fifty row, seventeen. Same company, but five more months on the benchmark you'd be compared against.

Elena ReyesSo ACV isn't a footnote, it picks your row.

Derek SimmonsAnd the row changes the budget you justify, which we'll get to.

Elena ReyesStress test. What if your customers span two rows?

Derek SimmonsThen you're really two businesses on one sheet. Put both rows in, label which customers go where, and don't average them into a number nobody actually sells at.

Elena ReyesOkay, but another wrinkle. Even the same report moves around year to year, right?

Derek SimmonsMarqeable summary of Benchmarkit shows the median going sixteen, then fourteen, then eighteen, then sixteen across calendar twenty-twenty-two to twenty-twenty-five.

Elena ReyesSo two months of wobble on the same yardstick.

Derek SimmonsWhich is why you pin the benchmark to its year. Write the year in the cell next to the number.

Elena ReyesDry, but I respect it. Cheap insurance.

Derek SimmonsCheapest line in the budget.

Elena ReyesSo land it. You're a million to three million in ARR. Which one do you use?

Derek SimmonsHighAlpha's band because it matches your size. KeyBanc's twenty months belongs to much larger companies. You're not them yet.

Elena ReyesEven KeyBanc isn't one number though, right?

Derek SimmonsThe survey also has a separate subgroup table with lower payback figures, but the sample as a whole still skews well above your size, so I wouldn't borrow either row.

Elena ReyesAnd Benchmarkit?

Derek SimmonsUseful only after you match your ACV. Before that, it's a ruler for somebody else's room.

Elena ReyesSo the twenty months isn't wrong, it's just not invited to your meeting.

Derek SimmonsPut that on a mug. Okay, so you've picked the benchmark that fits. Now comes the harder part, which is a target you can defend when your CFO asks where it came from.

Elena ReyesAnd when the board asks, because they will, where do we start?

Derek SimmonsWith motion. Self-serve and enterprise shouldn't share a number. Okay, so you've got a size-based range, and you've got a motion. Time to put actual cells in the spreadsheet. Formula first, and nobody gets to freestyle this one.

Elena ReyesFine, I'll be the skeptic. Read it out, and I'll poke it.

Derek SimmonsSales and marketing expense divided by new customer ARR times gross subscription margin, then times twelve. That's the version the Marqeable write-up gives.

Elena ReyesWhy the gross margin in there?

Derek SimmonsBecause a dollar of ARR isn't a dollar of cash. You're paying back out of margin. Classic problem: revenue payback flatters everybody.

Elena ReyesMm-hmm. And Marqeable is a secondary source, so treat it as the formula's address, not the final word.

Derek SimmonsAgreed. Worked example, made-up numbers. Say last year you spent six hundred thousand dollars on sales and marketing and landed five hundred thousand dollars in new customer ARR at eighty percent gross margin.

Elena ReyesLet me check it. Six hundred over five hundred is one point two. Divide by point eight, you get one point five. Times twelve, eighteen months. Hypothetical, but that's how you'd compute your own payback from last year's actuals.

Derek SimmonsNow the motion cross-check. Who's got targets by motion?

Elena ReyesMarqeable cites Bessemer's gross margin adjusted targets. Under twelve months for SMB, eighteen for mid-market? Yep, and twenty-four for enterprise.

Derek SimmonsWait for it. Enterprise at twenty-four.

Elena ReyesI see where you're going.

Derek SimmonsBecause step one, the HighAlpha band for a one to five million dollar ARR company said five to fourteen months. Twenty-four isn't even in the building.

Elena ReyesOkay, so let's stress test it in order. And everything we say from here is hypothetical, made-up inputs, not sourced figures.

Derek SimmonsStep one, you're a two million dollar ARR company, so you write down the five to fourteen band as your size reference. That's your first column.

Elena ReyesInvented numbers. Say it louder for the people in the back.

Derek SimmonsStep two, pick your motion row. Say you sell to small teams, self-serve with a little sales help. That's SMB. Bessemer's under twelve goes in the second column.

Elena ReyesAnd they agree, sort of. Twelve sits inside five to fourteen, cozy. Pick something around ten. Done. Go home?

Derek SimmonsCozy is the point. When both columns agree, the target is easy to defend.

Elena ReyesBoring. I want the fight.

Derek SimmonsFine. Step three, you're selling enterprise deals hypothetically to a handful of big accounts. Bessemer says twenty-four. Your size band tops out at fourteen.

Elena ReyesSo the sources conflict, and we say so out loud, and then somebody picks.

Derek SimmonsPick a number past the band's top, like eighteen, and write the trade-off next to it. A longer payback means you're financing each customer for longer.

Elena ReyesCash flow. Every extra month is a month your S&M dollars aren't back in the bank. At two million ARR, you probably don't have a cushion that fits a twenty-four-month clock.

Derek SimmonsOr the other direction, force it to the band's ceiling, fourteen, and you might starve the very deals that define an enterprise motion.

Elena ReyesWhich is a legit choice as long as it's written down. The defensible part is the note in the cell, not the number. And when the board asks why eighteen and not fourteen...

Derek SimmonsThen the note answers it. The size band stopped at fourteen, the motion target said twenty-four, and we chose eighteen knowing what it costs in cash.

Elena ReyesPin the year too. That benchmark wobble from earlier, put the source year next to every input.

Derek SimmonsAlready in the template, column D.

Elena ReyesOf course it is. So now you've got a target, ten, fourteen, eighteen, whatever you defended.

Derek SimmonsAnd a target by itself pays nobody. Last step, turn it into a twenty twenty-seven S&M budget. How many dollars does that clock actually let you spend?

Elena ReyesOh, now it gets expensive.

Derek SimmonsOkay, so we've got the target. Now flip the formula around and solve for the budget. Budget equals target payback in months times the new ARR you want times gross margin divided by twelve. Hypothetical numbers only. Nobody's real company here. Totally made up. Say you want one million in new ARR and gross margin is eighty percent. Fine, run the eight-month version first. Eight times one million is eight million times point eight is six point four million. Divide by twelve, you get about five hundred and thirty-three thousand dollars of S&M.

Elena ReyesHalf a million to buy a million in new ARR. That feels almost rude.

Derek SimmonsWait for it. Same inputs, twenty months.

Elena ReyesOh, no. I can feel it.

Derek SimmonsTwenty times one million times point eight over twelve. About one point three three million.

Elena ReyesWhoa, two and a half times the money, same goal.

Derek SimmonsSame spreadsheet, one cell changed.

Elena ReyesGive me a middle case too. Eight and twenty are the extremes.

Derek SimmonsFourteen months, same inputs. Fourteen times one million is fourteen million times point eight is eleven point two million. Divide by twelve, about nine hundred and thirty-three thousand. Right between the other two.

Elena ReyesWhich is why the target row isn't a cosmetic choice. Pick eight when your motion lives at twenty, and you've underfunded the plan by about eight hundred thousand dollars before January ends.

Derek SimmonsAnd you'll blame the sales team in March.

Elena ReyesClassic. Okay, but payback is only half of what drives growth, right?

Derek SimmonsRight. HighAlpha's twenty twenty-five Benchmarks Report pairs net revenue retention with payback as the main growth predictor.

Elena ReyesSo the model needs an NRR line sitting right next to payback.

Derek SimmonsYes. Otherwise, you're funding new customers and pretending the old ones stay put.

Elena ReyesHmm. What goes in that line, though?

Derek SimmonsYour own retention as an input you can change. Slide it up and down and watch how much new ARR you really need to hit the growth number.

Elena ReyesBecause if the base keeps more of its revenue, the new ARR goal shrinks and so does the budget. Try it. Hypothetically, take the fourteen-month case, but retention improves, so you only need eight hundred thousand in new ARR. Fourteen times point eight million times point eight over twelve is about seven hundred and forty-seven thousand dollars. So roughly a hundred and eighty-six thousand less from a line most people leave blank. Poke at it though, because the retention input is the soft one. It is. Retention is a forecast, not a fact. So I'd run the budget at your current retention and again at the improved one and fund the first.

Derek SimmonsFair. Now, the one thing we can't give you, burn multiple.

Elena ReyesEverybody asks for it.

Derek SimmonsOur sources don't cover it. No burn multiple benchmark came out of anything we pulled, so we're not quoting one.

Elena ReyesGood. A made-up benchmark is worse than a blank cell.

Derek SimmonsBlank cell with a label on it that says, "Go find one."

Elena ReyesSo the rule, choose your row.

Derek SimmonsPin the year of the source.

Elena ReyesSet the target from that row.

Derek SimmonsFund the budget with the formula.

Elena ReyesAnd recheck every quarter because the inputs move.

Derek SimmonsThat's the build. The wrap is next, and Elena has the spread to hand back.

Elena ReyesI do. Bring a pen.

Derek SimmonsSo one line, the spread is the finding. Pin the benchmark to your motion and your year, and the twenty twenty-seven budget stops being a guess.

Elena ReyesMm, and the guess was the expensive part.

Derek SimmonsAlways is.

Elena ReyesAnd it's a dull habit that fixes it. Pick your row, pin the year, and write down why. Nobody gets fired for a note in a cell.

Derek SimmonsRe-check it quarterly too. The inputs move, and so do the benchmarks.

Elena ReyesAnd if the year on your benchmark is stale, that's the first cell to fix.

Derek SimmonsStale benchmarks are how a budget quietly goes wrong.

Elena ReyesOkay, so ask time. If you know one founder who's staring at an eight and a twenty and picking whichever feels nicer, send them this episode today before their board deck.

Derek SimmonsDo it gently. Maybe don't text it with just a skull emoji.

Elena ReyesTempting though. Then subscribe, YouTube or whatever podcast app lives on your phone. Pick your poison.

Derek SimmonsAnd leave a review. It's how the show keeps finding founders willing to put real numbers on the table, and we run on those.

Elena ReyesReal numbers, not vibes.

Derek SimmonsNot even good vibes. Go fill in your blank cell, Elena.

Elena ReyesOh, I'm filling mine in tonight, Derek. Wish me luck with the pen.