Short answer. If your clients come back to you month after month, an app rarely brings in new revenue by itself. It pays for itself through staff time: the status calls, update calls and document chasing your team stops doing. So build the case on hours saved, priced at what those people cost you, over three to five years. If you can't name that work, you are not ready to build.
The maths in three steps
1. Count the manual work
List everyone who answers "where is my case?" calls, phones clients with updates, or moves documents between email, chat apps and your case system by hand. Oleksiy's rule of thumb: above 500 cases a month, that team is usually 10 to 15 people. Below 500 cases, he says, an app won't pay for itself, and a phone and a spreadsheet will do the job cheaper (his post of 1 July 2026).
2. Take the share the app removes
Since the Merezha Prava client app launched, the law firm has had 15% fewer inbound calls. Automated case-stage notifications cut the status-check calls: clients see where their case is before they pick up the phone. Documents now arrive in the app, not across Viber, WhatsApp, Telegram and email.
Where the freed time went matters as much as the percentage. In Oleksiy's words: "Same team, same desks - no one cut", and the information department "got their time back" (29 May and 14 May 2026). [TEAM SUPPLIES, optional, with the client's OK: how many people work in the information department, and how many hours a week the app gave back. Delete this sentence if not available.]
For your own team, assume a smaller drop than Merezha's. Allow for adoption, too. Moving tens of thousands of clients onto an app takes months, not weeks. Some clients can't use it at all, because of age, devices or digital skills (post of 14 May 2026).
3. Read it over years, from month 6
Multiply the saved hours by what each person costs you, fully loaded. Then read it across three to five years. The savings grow with your caseload, because you don't hire the next coordinator. Don't start the clock at launch, though. Our internal benchmark is that saved hours and fewer calls become countable around month 6. As Oleksiy puts it, that is "our internal benchmark, not a promise" (post of 30 July 2026).
How many hours a week must the app save?
Here is the same logic run backwards on our own cost model. We took the two default builds in our app cost calculator, which prices every hour at a blended planning rate of £45. Our actual rates are per engineer, about $30–90 an hour. We added 17.5% of the build cost each year for maintenance, the calculator's rule of thumb. Then we asked: how many staff hours a week must the app remove for the saved wages to cover all of it?
| First build (our cost calculator's defaults) | Build cost | Hours a week to break even over 3 years | Over 5 years |
|---|---|---|---|
| Simple first version, 589 hours | £26,505 (the calculator shows £22,500–£30,500) | about 10 | about 8 |
| Typical business app, 1,524 hours | £68,580 (the calculator shows £58,500–£79,000) | about 27 | about 20 |
How we worked it out: (build cost + years × 17.5% of build cost) ÷ years ÷ 52 weeks ÷ £25. The £25 is the loaded cost of one staff hour that our ROI calculator's legal template starts with. It is not a survey figure, so use your own. If savings only start around month 6, the 3-year figures rise by about a fifth and the 5-year figures by about a tenth.
How to read it: a typical business app pays for itself on labour alone if it takes about 20 hours a week off your team for five years. That is roughly half of one person's working week. If you can't point to that much repeated manual work, start with a smaller first version, or don't build yet.
What our own ROI calculator says
Our ROI calculator does count revenue. Its legal template adds up three lines: calls avoided, admin hours saved, and extra revenue from clients who come back. On the template's default settings (a £150,000 build, conservative scenario), staff time is about half of the modelled yearly benefit: £39,531 of £78,411. Returning-client revenue is the other half. Set the revenue line to zero, and payback moves from 23 months to about 46.
So the post's argument holds inside our own model. The labour half is the part you can measure and defend to a partner or a board. The revenue half is real, but it should not carry the case alone. The Merezha Prava breakdown has the full maths. It shows what one call costs, how many calls a month would repay a build on their own, and the calculator run line by line.
The rest is real, but it isn't the business case
Going digital, fewer anxious calls and a client community all matter. None of them shows up on an invoice, so none of them should carry the argument on its own.
If you can't name the work yet
Then wait with the decision. Oleksiy says he talks founders out of apps more often than into them (post of 1 July 2026). A discovery phase is where the work gets named: who does what by hand today, how often, and what it costs. If the work is there, our new app and MVP development starts from that list.
Run it on your numbers
Book a 30-minute ROI review call with Oleksiy, who wrote the post. Bring three numbers: how many cases you run a month, how many people answer client calls and move documents, and what one hour of their time costs you. Or try the ROI calculator first.
Related: 9 things 50+ law firm conversations taught us, Oleksiy's post of 13 August 2026 on how law firms talk to their clients.
The original post
Most client-heavy firms measure mobile app ROI as new revenue. Wrong line on the P&L. The real number is the LABOR you stop paying for.
In most cases an app is not a revenue source. It removes manual work you already pay salaries for. Think of it as a hire you no longer have to make.
Here is the math. Use your own numbers:
- Count the manual work. List the people who answer "where is my case?" calls and move documents by hand. Above 500 cases a month, the info team is usually 10 to 15 people.
- Take the share the app removes. For one of our clients, an app cut the info team's outbound calls by 15%. It also handled 6,700 document exchanges inside the app in the first months. For your own team, assume a smaller, conservative drop. That drop is hours your team stops spending.
- Read it over years. Multiply those saved hours by what each person costs you. Then read it across 3 to 5 years. The savings grow as your caseload grows. You stop hiring the next coordinator.
That labor line is the part you can measure.
The rest is real too: going digital, fewer anxious calls, a client community. They just do not show up on an invoice.
If you cannot name the work the app takes off your team, you are not ready to build.
Want me to run this on your real numbers? Link's in my profile.