Back

Essay

You Don't Need More Traffic. You Need More to Sell.

August 8, 20267 min read

Short answer: a widely cited SaaS benchmark puts the cost of a dollar of new annual contract value at roughly six times the cost of a dollar from an upsell or plan expansion. Every founder knows this. Almost every growth plan for a small software product is still an acquisition plan. If a few hundred engaged users aren't worth enough each, more of them won't fix it — and the fastest way to raise what each one is worth is usually not on your roadmap at all.

The arithmetic everyone agrees with and nobody acts on

Ask a founder how they plan to double revenue this year and you will hear about channels. Content, SEO, a launch, ads, a founder-led sales push, maybe a directory or two. All of it is the same shape: get more people to the same offer.

Now ask the same founder whether it is cheaper to sell to a stranger or to a customer who already pays them. Nobody gets that wrong. The benchmarks are lopsided — a dollar of new ACV costs multiples of a dollar of expansion, because expansion skips the entire cost of establishing trust.

So there is no disagreement about the economics. The gap is elsewhere: acquisition is a plan you can start on Monday, and expansion isn't. Expansion needs something new to sell, and "something new to sell" reads as roadmap — quarters of engineering, an unvalidated bet, opportunity cost. Faced with a project you can start tomorrow and a project that needs a quarter, founders reliably pick the one they can start tomorrow, even when they know its return is worse.

Why the acquisition plan is getting worse, not better

There is a second reason to care, and it has changed recently.

Building software got easy. That is not a complaint and it is not nostalgia — it is the single most important fact about the market you are selling into. The consequence is that the number of competent products chasing your customer's attention went up sharply, while the amount of attention stayed exactly the same. Every channel that worked in 2022 now has ten times the supply pushing through it.

Which means the acquisition plan is not just the more expensive option. It is the option whose price rises every quarter, because you are bidding against a growing number of products for a fixed amount of attention.

And here is the turn, because it is the whole argument in one move: those products competing with you for your user's attention already have your user. The crowd you are trying to outbid is a crowd of tools sitting inside the same customer's workflow. Outbidding them for attention is the most expensive possible use of that fact.

Three levers, and the one that's actually available

Revenue per account moves through exactly three things: price, packaging, and how much there is to buy.

  • Price. Real, immediate, and mostly a one-time move. You can raise prices once a year without a fight; you cannot raise them quarterly.
  • Packaging. Also real — removing a tier nobody should be on, or moving a feature up a level, can lift ARPU without touching the product. Also finite. There are only so many times you can rearrange the same value.
  • More worth buying. Unbounded, and the one everyone treats as a roadmap problem.

The third lever is where the room is, and the reason it feels unavailable is an assumption buried in it: that anything you sell has to be something you built.

The version that doesn't need a roadmap

Your users are already paying someone else to finish the job your product starts. Not hypothetically — go and look at what they do in the ten minutes after they close your app. Somewhere in that sequence is a step they leave your product to complete, and a company charging them for it.

A cross-sell integration is what happens when that capability moves inside your product instead. The partner's feature runs in your app, your user finishes the job without a second login or a second subscription, you charge for it as your own add-on or tier, and a signed agreement routes an agreed share of that revenue to the partner automatically.

Look at what that does to each objection to lever three:

  • It isn't a quarter of engineering. You are connecting to something that already exists and already works. That is usually days.
  • It isn't an unvalidated bet. Your users are already paying for this outcome, just to someone else. That is stronger demand evidence than any survey you could run, and it is why "integrate first, build later if it earns" is a cheaper way to validate a feature than building it.
  • It doesn't need traffic. The revenue comes from accounts you already have, which is exactly what makes it expansion rather than acquisition.

And it runs in both directions. In a group of complementary products, each member is the host in one integration and the partner in the next — so the same relationship raises everyone's revenue per account rather than moving revenue from one member to another.

What this is not

Two things, because both get read into it immediately.

It is not a referral. Sending your user somewhere else moves a lead and pays you once; it also makes your product a smaller part of their day. An integration changes what you are able to sell — your users can do something inside your app they couldn't before, and you charge for it. The direction of the user matters: in one case they leave, in the other they stay and buy more.

It is not a reach play. The temptation is to sell this as audience multiplication: five partners, five audiences, five times the exposure. That claim collapses the moment a founder does the arithmetic on a product with 300 users, and it deserves to. The honest version is smaller and holds at any size: the same audience hears from five of you instead of one of you, and each of you has more to sell them.

How to tell whether this is your problem

Not everyone should stop working on acquisition. Two questions separate them.

Are your existing users engaged? If people use the product weekly and stay, you have earned trust and are under-monetising it. If they churn in month two, expansion is the wrong project — you have a retention problem, and adding things to sell to people who are leaving is a way to make less of both.

Do they leave to finish the job? If you can name the step and the tool they use for it, you have found the integration. If your product genuinely completes the job on its own, there may be nothing adjacent to sell, and pricing and packaging are your only levers.

Engaged users plus an obvious adjacent step is the profile where this works, and it is a very common profile for a product with a few hundred users and a distribution problem.

The honest part

The coordination is the work: finding a partner whose users are genuinely your users, agreeing what each side ships, papering it so an outage has an owner, and settling the money every month without a spreadsheet argument. That is what Ordana automates — matching, the integration plan, the contract with identity-verified signatures, revenue-share rules scoped to specific Stripe products, and automatic payouts on every in-scope charge, out of each member's own Stripe account. Money is never pooled and each member keeps their own pricing.

Ordana does not write the integration. The generated prompt, the SDK and the revenue tag are unbuilt; today the members write the code themselves. It is a few days of work rather than a quarter, which is the point — but it is not zero, and pretending otherwise would be the kind of promise that costs trust the first time someone signs up.

The takeaway

The reflex when growth stalls is to go and find more people. It is the reflex because it is the thing you can start immediately, not because it is the thing that works — and it gets more expensive every quarter as the market fills up with products competing for the same attention.

The cheaper move is to ask what else the users you already have would pay for, and then to notice that the answer is usually already built, by a company standing one step further along the same job. You do not need more traffic. You need more to sell to the traffic you already have.

Related reading:

See how cross-sell integrations work on Ordana →