The Taxi Apps Can Take the Work. They Shouldn't Take the Customer

For licensed London Hackney drivers. UK-only.

When the taxi apps first arrived, the deal looked simple: they brought the work, and in return they sat between you and your customer. Fair enough, at first.

But once an app owns the customer relationship, the balance quietly shifts. The drivers who fit what the app wants see the best work first; everyone else takes what's left. When the app says jump, the ones who ask "How high?" get the jobs. That's not bad luck and it's not personal — it's just how the model is built to work.

How did it get to this? In large part, the courts. Once the apps won the key legal battles (the detail's below), the door stayed open — and that's exactly when they started taking real control of the customer relationship.

So here's how it actually works:

Every time you take an app job, the customer's relationship is with the app. Not you. They paid the app. The app paid you. The customer's phone has the app's number, not yours. Next time they need a cab, they open the app. The app sends them whoever's nearest — which might be you, but probably isn't. The work was yours for that one fare. The customer was the app's the whole way through — which is now truer than ever.

The Hackney trade has known this for a decade. The mass black-cab protests of 2014 and 2016, the LTDA's private prosecution that became Transport for London v Uber London Ltd [2015] EWHC 2918 (Admin), the LTDA-LPHCA joint opposition that argued smartphones used as fare-calculators were taximeters under the Private Hire Vehicles (London) Act 1998 — none of that succeeded in stopping the app-mediated customer-relationship transfer. Mr Justice Ouseley ruled in October 2015 that smartphones aren't taximeters, the legal door stayed open, and PHVs operating via apps have been picking up off the street like immediate-hire ever since — despite being legally pre-book-only. The Hackney monopoly on ply-for-hire wasn't abolished; it was hollowed out. The trade has been watching the customer relationship walk into apps — and not just into Hackney apps; into PHV-via-apps too — for the entire decade since.

This post is about doing something about that — without quitting the apps, without picking another fight the trade body has already lost, and without losing the income those apps reliably provide.

The honest situation

Let's name where we actually are.

Most London cabbies use one of the apps at least some of the time. There's a big one for the Hackney trade specifically. Cabbies use it because it's there, the work is steady, and on a quiet afternoon a Heathrow job from the app is better than no Heathrow job. Nobody reading this is going to be talked out of using the apps entirely. The apps work — for the apps.

What the apps don't do — what they structurally can't do, because their business model depends on it — is give you the customer's contact details. Every app fare you complete results in:

The fare paid (good for you, in the short term).
The customer record retained by the app (good for the app, forever).
Zero way for the customer to contact you directly next time (good for the app, neutral or bad for you).

This is the structural lock-in, and it isn't a bug. It's how the apps work. They get paid because they own the customer relationship. If they handed the customer your contact details at the end of every job, they'd be undermining the asset they exist to monetise.

So the apps aren't going to change. The question is what you do.

The "do what the app wants" thing, briefly

Just to put it in its proper place: yes, the apps reward the drivers who do exactly what the app wants, and the ones who don't tend to see less work — sometimes none, when it's quiet. That's real, and it's genuinely tough on the drivers it happens to.

But the more interesting version of the story is this: the cabbie who's been quietly building a direct customer book on the side would barely notice if an app cut their work tomorrow. Because that cabbie's income wasn't really coming from the app anyway — it was coming from the regulars who book them direct.

Deactivation insurance isn't an appeal process. It's a customer book that's yours.

Co-existence with the apps — for real

The framing that lands with most cabbies in the 16-strong network the founder has been operating with for the last 18 months is this:

The apps can have the work. They shouldn't own the customer relationship.

You take the app fare. The work is the work. The fare is fine. You'd be silly to turn it down on principle when your meter could be running.

But at the end of that fare — the thirty seconds before the customer climbs out of the cab — there's a window where the relationship can become yours, not the app's. That window is what most cabbies don't use, because nobody has ever explained that it's there. Every card you hand over at the kerb tips a little more of that customer relationship back into your hands — and the more of it you own, the more in control of your own work you are.

The mechanic is the same one as the airport conversion loop. At the kerb, you hand the passenger your card (a business card, or — if you're a London Driver Pack cabbie — a printed QR card from the Driver Pack). You say one line:

"Next time you need a cab, message me direct — I'll do you a fixed price the same way."

That's it. No upsell. No leaflet. No "and by the way, the app charges X%". One sentence at the kerb.

A meaningful fraction of passengers take the card. Of those, a meaningful fraction use it. Of those, a meaningful fraction become regulars. The work paid you for that fare. The conversion compounds for years.

The app got the fare. You got the customer. Everyone's happy except the structural lock-in.

Why this is a different category of solution

Most cabbies who try to build a direct customer book hit one of three walls:

The compliance wall. Taking a pre-booked job from a regular off WhatsApp without a proper booking record puts the badge at risk. (Post #4 covers this in detail.) Solution: BAT.TAXI's invite system generates a proper TfL-compliant pre-booking record at the moment the customer confirms a quote.
The admin wall. Keeping track of who booked what, when, at what price, with what notes — in a notebook or a personal spreadsheet — is fine for a week and fails by the second month. (Post #3 covers the proper CRM behind this.) Solution: BAT.TAXI's customer record is the proper version of what cabbies have been keeping in their head and their phone for years.
The "what happens when I'm not available" wall. A regular books you for the same week you're on holiday. Solution: the Co Driver mechanic — when you can't cover one of your own bookings, you choose to offer it out to a trusted Co Driver, completed on your terms (same price, same conditions). The customer experience is identical regardless of which driver fulfils. (Post #4 covers the mechanic.)

All three walls are why the manual version of the "convert at the kerb" approach has historically been hard to scale. BAT.TAXI is the infrastructure that lets a cabbie do this systematically — without losing the licence, the time, or the customer to a clash.

The customer-cascade-deletion proof point

If you take only one structural fact from this post, take this one:

BAT.TAXI cascade-deletes your customer list if you ever leave. Your personal customer list — every contact and every booking tied to you — is deleted from our database and is unrecoverable. (We keep only the minimal information the law requires — never used commercially, and only ever disclosed on an official, documented request from the licensing authority or law enforcement — none of which can rebuild your customer list.)

And while you're with us, those customers are genuinely yours. Every customer you sign up gets their own app to book you, linked only to you — no other driver is ever connected to your customers. The passenger app is free to them. You can add unlimited customers for one flat fee, whether that's one regular or a thousand — the price never moves.

That's the test of whether a platform really treats your customers as yours. The apps don't cascade-delete because the customers were never yours. BAT.TAXI does because they are.

This is the structural opposite of the lock-in problem we started with. The lock-in problem says "the customer's relationship is with the platform, not the cabbie." The cascade deletion says "prove it — try to leave, and watch what walks out with you."

The Ark logic, applied to apps

This is the same logic the Waymo post made for the driverless future, applied one step earlier. And it's worth knowing why the apps will never hand the customer relationship back on their own: the companies that own and back them are, in many cases, the same ones building the driverless cars (who really owns the taxi apps). The Ark — your direct customer book — is built one passenger at a time. Every app fare, every rank fare, every casual hail is a chance to add one passenger to it.

The cabbies who do this systematically across 2026 arrive at:

Q4 2026 (Waymo launches): with a customer book the robotaxis structurally can't take.
The next time an app changes its commission structure: with a customer book that doesn't care.
The next time an app suspends or deactivates them: with a customer book that keeps earning.

The Ark isn't built for one specific catastrophe. It's built because the cabbie who owns the customer relationship is structurally protected against whatever comes next. Deactivation. Driverless. Surge changes. Commission hikes. New apps entering the market. Whatever it is — your regulars still have your number. They still message you direct. You're insulated.

What to do tonight, before tomorrow's shift

Three things, in order of importance:

Join the London Taxi pre-launch waiting list. bat.taxi/driver-location-selection. £4.99 a month - cancel anytime (Platform Only) or £99 first year for the Driver Pack (London only — includes physical in-cab QR materials + Airport Pricing dashboard feature; platform renews at £39.99/year from Year 2).
Decide on your kerb-side line. "Next time you need a cab, message me direct — I'll do you a fixed price the same way." Or your own variant. Rehearse it until it's natural. The first ten times you say it out loud it'll feel awkward. By the twentieth it's reflex.
Commit to one passenger added per shift. Not three. Not five. Not "convert everyone". Just one. Compound it through the year. By Christmas you'll have a customer book that's structurally yours — not the apps', not BAT.TAXI's, not anyone's. Yours.

FAQ

Should I quit the apps? No. Quitting cold is the wrong move. The apps still pay you for actual work, and most cabbies in the network the founder has been operating with use the apps as a supplemental income while building the direct book on the side. Read the 90-day plan for how the apps' share of your income naturally shrinks over time as the direct book grows — without you having to choose a side.

How long before this actually replaces app income? The 16-cabbie network the founder has been operating with for 18 months saw direct work overtake app work as a share of weekly hours somewhere between months 6 and 12, depending on how disciplined each cabbie was about the kerb-side conversion. Your mileage will vary. The 90-day plan above sets a realistic curve.

What if I get deactivated from an app while building this? Then the work that was going through that app stops, and the customer book you've been building keeps going. The cabbies in the BAT.TAXI network with the largest direct books are the least affected by app deactivations. The whole point of the Ark logic is that it's the deactivation insurance — not an appeal process.

Is the apps' commission really 29% on average? According to Oxford / Worker Info Exchange research published in June 2025, the average take rate on the leading UK ride-hail platform is around 29%, with some cases exceeding 50%. UK drivers across the major platforms collectively lost an estimated $1.6bn between April 2024 and March 2025 to algorithmic pricing changes. This is for context — the chronic customer-ownership problem this post is about exists regardless of commission rates.

What about the rank? Same logic. Every rank fare is recruitment time. The passenger getting in the back is, by definition, someone who needed a cab today and may need one again next week. Hand them the card at the end of the journey. The mechanic is identical to the app version.

Is BAT.TAXI poaching customers from the apps? No. The customer paid the app for that specific journey, and the app got its fare. What BAT.TAXI helps you do is invite the customer to book direct with you next time — which is something the customer has always had the right to do. The cabbie is just making that future booking easier and TfL-compliant. The app isn't being poached from; the future business is being earned. If the taxi apps say drivers are poaching, drivers might say that's why they have all our customers (pot-kettle analogy), the fact of the matter is, passengers will use services best suited to their needs.

Has anyone actually done this? Yes. The founder of BAT.TAXI, Peter Schive, plus fifteen other London cabbies, have been operating this exact model manually for the last eighteen months — physical in-cab marketing, business cards, WhatsApp overflow between the network. The BAT.TAXI platform is the digitalisation of that working service. See post #1 and the founder bio.

Ready to keep doing app work — but stop losing the customers? Join the London Taxi pre-launch waiting list →

Already a BAT.TAXI cabbie? Sign in →