I think a lot of growth work gets framed as persuasion when it is really an exchange.
We want the user to invite a teammate.
Start a project.
Connect a data source.
Come back tomorrow.
Turn notifications on.
Give us one more step, one more click, one more bit of trust.
Sometimes those asks are completely reasonable. Sometimes they are the exact move that helps the user get to value.
But the ask is never free.
Every growth surface is proposing a trade.
The user gives up a little time, attention, certainty, social capital, or setup effort. In return, they expect something. Maybe speed. Maybe clarity. Maybe progress. Maybe relief. Maybe the feeling that this product is finally going to do the annoying thing they came here to solve.
When teams name that trade clearly, the product tends to get better. When they do not, they usually end up optimizing pressure.
That is how you get prompts that technically lift conversion while leaving users a little less sure they made the right choice.
The job is not only to increase the yes rate
This is one reason I keep coming back to interaction cost as a useful growth concept, not only a UX one.
Raluca Budiu describes interaction cost as the effort users spend to achieve their goals on a site or app. That sounds simple, but I think it explains a lot of shaky growth work.
A team sees a desirable behavior and asks how to get more of it.
More clicks on the CTA.
More onboarding completion.
More collaboration invites.
More users returning in week one.
All fair questions.
But sometimes the better question is whether the product has made the trade worth taking.
If a user has to read too much, guess too much, wait too long, grant too much access, or do too much setup before the payoff appears, the product is asking for an expensive trade.
And expensive trades do not disappear just because the button copy got sharper.
They become hidden debt.
Users are hiring progress, not ceremonies
I also think this is why Jobs to Be Done still holds up.
People do not arrive wanting to complete your internal milestone taxonomy. They are trying to make progress in their own context.
That matters because the same product ask can feel cheap or expensive depending on what progress the user thinks it unlocks.
A teammate invite feels reasonable if the user has just finished a workflow and can now make the work collaborative.
The same invite can feel premature if the user is still trying to answer a more basic question.
Does this thing even help me.
That is the part growth teams can forget when they inherit a funnel and start polishing the steps.
The user is not reacting to the ask in isolation.
They are reacting to the full trade.
What am I being asked to give.
What do I get back.
How soon do I get it.
How certain am I that the payoff is real.
What happens if I am wrong.
That is a more useful lens than whether the event name sits in the right stage of the dashboard.
A lot of bad nudges are bad trade design
I do not think most weak onboarding and lifecycle prompts fail because the team forgot persuasion principles.
Usually they fail because the product is trying to collect payment before it has earned trust.
You see it in all kinds of places.
A signup flow asks for a long setup ritual before showing evidence the tool works.
A product pushes for notifications before the user has experienced a single worthwhile reminder.
A collaboration product asks for invites before the first user has created anything worth sharing.
A lifecycle email drives someone back to a generic homepage instead of the unfinished work they actually care about.
An SEO landing page makes a bold promise, then hands the visitor a blank state and a scavenger hunt.
None of those are messaging problems first.
They are exchange problems.
The product is asking for something costly while returning something vague.
That is rarely a good trade.
This is where backstage work matters
I like service blueprints for the same reason.
Sarah Gibbons describes them as a way to visualize the people, processes, and evidence tied to a customer journey. That is useful because a lot of product asks depend on backstage conditions the user never sees directly.
A prompt only feels smart if the timing is credible.
An integration request only feels worth it if the import works.
A return email only feels helpful if the deep link lands in a coherent state.
A collaborative ask only feels natural if permissions, empty states, and follow-up notifications do not immediately create confusion.
Sometimes a team keeps pushing harder on the visible surface when the real problem lives one layer underneath.
The trade is failing because the promised return is operationally shaky.
That is not a copy issue.
That is a systems issue wearing a growth costume.
The artifact I like is a trade note
When a team is stuck debating whether a prompt is too early, whether onboarding is too heavy, or why a lifecycle nudge gets clicks but weak downstream behavior, I would write a trade note.
It is small on purpose.
Not a giant strategy deck.
Not a perfect framework with ten swimlanes.
Just one page that forces the exchange into plain view.
Trade note
- The user moment
- The behavior we are asking for
- What the user has to spend to do it
- What payoff they expect in return
- How quickly that payoff becomes visible
- What proof makes the ask feel safe
- What happens if they choose not yet
- What backstage dependency could break the promise
- What downstream behavior would tell us the trade was actually good
That last line matters a lot.
A trade is not good because the user said yes once.
A trade is good when the user says yes and the experience that follows justifies the decision.
Otherwise the product is borrowing trust and spending it badly.
What changes when you work this way
First, you get less impressed by raw conversion lifts.
If a prompt goes up, you still ask whether the trade improved or whether the pressure increased. That usually leads to better guardrails and less self deception.
Second, onboarding decisions get cleaner.
You stop asking whether a step is important in the abstract and start asking whether this is the right moment to charge the user for it.
Those are different questions.
Some steps are absolutely important. They still may be poorly timed.
Third, lifecycle work gets more honest.
A reminder is not useful because it resurrected a click. It is useful if it brings the user back to something legible, valuable, and easy to resume.
Fourth, acquisition and product start talking like they belong to the same company.
If search, landing pages, ads, sales, onboarding, and lifecycle are all participating in the same trade, then the handoff quality matters more than any one local conversion surface.
That is especially true in growth roles, where the easiest way to create fake momentum is to let one part of the journey overpromise while another part quietly absorbs the damage.
A simple gut check
Whenever a team wants more of a behavior, I think it is worth asking a few blunt questions.
- What is the user paying here
- Is the return specific or hand-wavy
- How long do they wait before they feel repaid
- If they say yes, does the next screen prove they were right
- If they say no, have we learned something real about readiness
If those answers are mushy, the growth problem is probably not solved yet.
You may still ship the test. You may still need the prompt. But at least you will know what kind of bet you are actually making.
The point
I do not think the craft of growth product is mainly about getting better at asking.
It is about getting better at structuring the exchange.
Good teams learn how to make the next step feel proportionate, credible, and worth it.
They know when to ask for less.
They know when to show proof earlier.
They know when to move a prompt later.
They know when the copy is fine and the trade itself is the problem.
That discipline sounds softer than funnel optimization, but I think it is actually more rigorous.
It forces you to account for user effort, user trust, operational reality, and downstream quality all at once.
Which is another way of saying it forces you to practice product judgment.
And in growth work, product judgment is usually the difference between a behavior lift you can trust and one you have to explain away later.
