Packaging and pricing: what you charge per, what you sell, then how much.
Three decisions in strict order, and no invented numbers: where evidence exists the number is stated; where it does not, you get the test that produces one, formatted and ready to run.
Costs you an adviser's hour, every time. Promised unlimited, free.
The employer product
Costs you an adviser's hour, every time. No price anywhere on the page.
Real cost per customer No real cost
0people have been asked what they would pay. You recorded price against three lost deals.
You are giving away the only part of the product that costs money to deliver.
The full output tests four candidate metrics, designs the packages for both motions, costs everything you give away, and sets out the method that produces the number.
Built on WealthKarma, a retirement app for Gulf expats. Every count is real and every gap is marked as a gap.
Packaging and pricing0 of 4 numbers tested · 12 deals, 17 interviews, employers page · 19 Aug 2026
01 · What the alternatives charge
Doing nothing
Free. Chosen by 5 of 12.
Someone they knew
Paid by commission on what they sell you. The buyer never sees a fee. Chosen by 4 of 12.
The employer scheme
Free, and already happening. Chosen by 3 of 12.
The four apps
All publish a headline fee. Numbers not read for this outputBenchmark
Three of the four alternatives cost the buyer nothing at the moment they choose. Whatever you charge is being compared against zero, not against Sarwa.
02 · What you blamed, and what they said
You wrote "too expensive"
Buyers who named price
You recorded price three times as often as buyers raised it. Your pricing anxiety is not currently evidence, and it should not set the number.
03 · The metric
What you charge per. Structural, and the hardest thing here to change later.
Direct
Candidate
Align
Pred
Def
Meas
Known
Verdict
A share of what they've saved
Fails
A flat monthly fee
Take this
Per adviser call
Add-on
The share-of-savings metric passes every test and still fails. It is what all four competitors use, and it earns nothing from someone who hasn't started saving. That is precisely the person the positioning targets. The metric and the position cannot both be right, and the position has more evidence behind it.
A flat fee is weaker on alignment — it doesn't grow as they save. Accept that. It is the only metric that earns from day one from somebody with nothing saved, and it is listed in your own table-stakes read as the one thing no competitor does.
Through an employer
Per employee per month. Standard for a benefit, predictable for the finance seat, and it is the only structure that answers the question that seat is currently asking into silence.
04 · The packages
Direct · FreeThe engine
Who it's for
Anyone who has been meaning to start.
What's in it
The levelled two-minute lessons. The wellness check that names the gaps.
Costs you
Near nothing per user. This is acquisition, not a loss leader.
Direct · PaidFlat monthly
Who it's for
People who finished the lessons and want someone to check their plan.
Forces the upgrade
Access to an adviser. It is the only thing here with real cost behind it, so it is the only honest gate.
Price
Not knownSee 06
EmployerPer employee, per month
What's in it
Everything in the free tier for every employee, plus a capped number of adviser calls.
Price
Not knownSee 06
What you refuse to sell
The simulator, insurance, estate and tax planning, and decumulation advice. Nobody has asked for any of them. They do not become tiers until somebody does.
05 · What you're giving away
"Unlimited 1:1 access to financial advisors"
On the employers page, for every employee, at no published price. An adviser hour has a real cost and there is no ceiling on this promise. Cap it, price it, or move it behind the paid tier before another employer signs.
The lessons and the wellness check
Near-zero marginal cost. Keep them free. This is the one thing you should be giving away.
06 · The number
No willingness-to-pay research exists. Nothing in this document tells you what to charge, and any figure written here would be invented.
The test, in order
First, free
Add four questions to interviews you are already running. What do you pay today for anything like this. What would you have to give up to pay for it. What would be so cheap you'd doubt it. What would be so much you'd stop.
Then, the range
Same four questions to 30 to 40 people. Plot them. Report the range where too cheap and too expensive cross. Never report a midpoint on its own.
Employer side
Ask the one company already using it what they would have paid, and what they compared it to.
Then decide
Take the top of the range. Going lower needs a written reason. Underpricing is the most common regret in this category and the fear of charging too much is not supported by what happens to companies that do.
A flat $5 a month
Appears in your own table-stakes read as an attribute. It is an internal idea, not a finding. Nobody has been asked.
Per employee, per month
No figure exists. The finance seat has been asking this into silence since the page went up.
Tested with buyers Nobody has asked Something of yours disagrees
07 · What would prove this wrong
The test
Kill it if
Fewer than half of twenty people asked can name any monthly figure they would pay.
Checked by
30 November 2026.
Why that
A flat fee only works if people can picture paying one. If they cannot name a number, they are not going to pay it, and the free tier is the whole business.
Goes stale if
Any of the four moves off a share of savings onto a flat fee. That is the gap you are stepping into.
Owner
Unassigned. One name, not a committee. Most companies never do this and it is why pricing drifts.
The finding
You are giving away the only part of the product that costs money to deliver.
08 · What can't be known
No willingness-to-pay research exists. Not one number in this document has been tested with a buyer.
Competitor fee levels were not read for this output. The table-stakes read establishes only that all four publish one.
One company uses the employer product and has never been asked what it would have paid.
Twelve deals is a pattern, not a statistic, and only one buyer in it mentioned price at all.
Where it sits
Research
Compete
Position
Message
Launch
Measure
you are here · 15 of 34
Want this for your company
This is one of thirty-four. The diagnostic tells you which ones you're missing.