This guide walks through the anatomy of a modern taxi dispatch stack, the real cost math, the integrations that matter, the compliance layer regulators care about, and a 90-day rollout you can execute without missing a single booking.
Chapter 1 · What Dispatch Software Actually Is in 2026
A generation ago, "dispatch" meant a room, a base radio and a whiteboard. In 2026 it means a piece of software that takes a booking from any channel — phone, app, WhatsApp, corporate portal — assigns it to the right vehicle in under a second, tracks the trip, charges the card, and produces a compliance-ready record. Everything else (marketing, fleet ops, HR, accounting) hangs off that spine.
Chapter 2 · The Five-Layer Stack
Live map, queue, auto-dispatch rules, manual override.
Job accept, navigation, cash/card, offline queueing.
Branded app + web widget + phone-line fallback.
Stripe/Square, corporate on-account, PDF invoices.
TfL/DVSA/TLC-ready records, retention, exports.
Chapter 3 · How Auto-Dispatch Really Works
Modern auto-dispatch is a scoring engine, not a lottery. Every available driver is ranked on: distance to pickup, expected time of arrival, vehicle class match, historical acceptance rate, and current shift earnings balance. The top-ranked driver gets a 15–30 second offer window; if they decline or time out, the offer cascades. The dispatcher can override any decision — the point is that in 95% of cases they don't have to.
Chapter 4 · The Booking Channels You Must Support
- Native passenger app — iOS and Android under your own developer accounts. Push notifications drive 3–4× the repeat booking rate of web-only.
- Web booking widget — embedded on your marketing site; captures organic search traffic without forcing an install.
- Corporate portal — cost centres, PO fields, group booking, monthly invoice.
- Phone line — because 30–45% of taxi bookings in the UK are still voice-first. A CTI plug-in that pops the caller record on screen is table stakes.
- WhatsApp / SMS — a one-line booking channel for the customers who refuse to install anything.
Chapter 5 · Pricing Models Compared
| Model | Typical cost | Best for | Risk |
|---|---|---|---|
| Ride-hail marketplace | 15–25% commission | 1–3 vehicles testing market | You rent your customer base |
| Entry SaaS | £79–£159/mo + £15/driver | 5–15 vehicles | Feature ceiling hits fast |
| Mid SaaS | £99–£249/mo + seats + 0–2% | 10–40 vehicles | TCO creep from add-ons |
| Enterprise | £1,500–£5,000/mo | 80+ vehicles | Multi-year lock-in |
| Owned white-label | £6.5k–£8.5k one-time + hosting | Any fleet 5+ vehicles serious about brand | Requires a proper migration |
Chapter 6 · The Integrations That Move the Needle
- Google Maps / Mapbox — routing, ETAs, place autocomplete. Non-negotiable.
- Stripe / Square / SumUp — direct-to-your-bank card processing.
- Twilio / Vonage — SMS fallback for booking confirmations and driver arrivals.
- Flight status API (FlightAware, AeroDataBox) — for airport jobs. Delayed pickups without this cost you £10–£30 per missed slot.
- Xero / QuickBooks / Sage — accounting export for corporate invoicing.
- What3words — pinpoint pickup at events, festivals, private residences.
- Open API — corporate clients increasingly demand a bookings endpoint they can hit from their travel-management system.
Chapter 7 · The Compliance Layer Regulators Actually Check
Every jurisdiction has its own retention window, but the audit fields are broadly the same: booking timestamp, pickup and drop-off coordinates, driver ID and licence, vehicle plate and licence, fare charged, payment method, complaint flag. Your platform must store all seven for the local retention period (UK: 6 months minimum, TfL specifies 12; New York TLC: 3 years; California CPUC: 4 years). If you can't export this in CSV in under 60 seconds, you have the wrong platform.
Chapter 8 · The Metrics That Matter Weekly
- Acceptance rate — % of auto-dispatched jobs accepted first-offer. Target >85%.
- Time-to-pickup — median from booking to on-scene. Target <8 min urban.
- Cancellation rate — passenger + driver cancels combined. Target <7%.
- Repeat-booking rate — % of unique passengers who book again within 30 days. Target >35%.
- Corporate share of revenue — target 25–40% for a resilient P&L.
Chapter 9 · The 90-Day Rollout Plan
Days 1–30 · Configure & submit. Brand assets, tariff table, driver list, zones, airport pricing, corporate accounts imported. Apps submitted to Apple + Google (allow 5–10 business days review).
Days 31–60 · Parallel run & train. New platform live for a single shift alongside the old dispatcher. Full driver training. Corporate clients notified with a two-week switchover window.
Days 61–90 · Cutover & stabilise. Old dispatcher read-only. Booking widget replaced on your website. Weekly metrics review; tune auto-dispatch scoring on real data.
Chapter 10 · Where Operators Lose Money in Year One
- Signing a per-trip commission deal to "avoid up-front cost". Pays back in fees within 8 months.
- Letting the vendor keep the Apple + Google developer accounts. You lose the apps on churn.
- Skipping corporate account setup — the single highest-margin channel in most fleets.
- Under-training dispatchers on manual override. Auto-dispatch works in 95% of cases; the 5% is what saves the shift.
- No airport flight tracking. You will miss 10–20% of airport pickups without it.
Bottom Line
Dispatch software is not a line item — it is the operating system of the business. Choose it on total 3-year cost, brand ownership, compliance fit and the integration surface, not on the sticker price of month one. Done right, a modern stack pays for itself in the first quarter and outlasts three cycles of vehicles.
Further reading: Best dispatch software 2026 · Small-fleet buyer's blueprint · Zero-commission dispatch · Enterprise pricing.
