
Costs and pricing
Part of Budgeting for marketing attribution across cash, accounting and economic views
Comparing attribution pricing structures on identical fields, values left blank
Compare attribution pricing structures in England using common commercial fields, buyer exposure and deliberately blank evidence-led values.
What to take away
- Most enterprise attribution suites publish no list price, so a fair comparison starts with the meter.
- Google Analytics 4 standard is the one named vendor with a published entry price of zero.
- GA4 360, Adobe Analytics, AppsFlyer and Adjust sell by quote, with the unit set in negotiation.
- Subscription, seat, usage, module, managed service and outcome-linked deals shift risk between the two parties.
- Check counting rules, timezone and duplicate handling before you accept any quoted volume.
What each named vendor publishes
These seven vendors cover the pricing structures in this comparison. Four of them publish nothing.
| Vendor | Main charge basis | Price position |
|---|---|---|
| Google Analytics 4, standard | Property and event collection | Free to use |
| Google Analytics 4, 360 | Annual contract with event tiers | Quote only |
| Adobe Analytics | Report suites and monthly unique visitors | Quote only |
| Mixpanel | Monthly events and data retention | Free entry tier, then paid tiers |
| Amplitude | Monthly tracked users and events | Free entry tier, then paid tiers |
| AppsFlyer | Attributed installs or monthly active users | Quote only |
| Adjust | Attributed installs or monthly active users | Quote only |
Quote-only means the vendor sets the unit, the minimum commitment and the discount inside a sales process. Public positions change, so confirm each figure on the vendor's pricing page on the day you compare.
Six structures appear across these quotes: subscription, seat, usage, module, managed service and outcome-linked, each placing risk differently.
Leave price, volume, service level, weight and threshold blank only where no written quote or source record exists. One blank field beats a blank column, because a missing figure invites a guess.
Meter tests before you accept a quote
- Define the exact event, record, user, connector or run that the supplier counts.
- Ask which timezone closes the billing period and when the count is taken.
- Ask how retries, duplicates, bots and late arrivals are treated.
- Ask whether deleted or suppressed records still contribute to the invoice.
- Request an export that reproduces the invoice quantity line by line.
If the meter cannot be audited, its cost exposure stays unresolved. Our budget template puts an evidence field on every money row, so a disputed quantity is easier to argue.
Example: two quotes on one grid
The grid below puts two illustrative quotes on identical fields. Every figure is a labelled placeholder entered by the buyer for modelling, not a vendor price.
Two quotes on one grid
Quote A (illustrative)
- Charge basis
- Events at ingest
- Unit
- One event per request
- Included volume
- 2,000,000 events a month
- Minimum term
- 12 months
- Placeholder cost
- £1,200 a month
- Exposure if usage doubles
- Cost rises with events
Quote B (illustrative)
- Charge basis
- Named seats
- Unit
- One active user per month
- Included volume
- 25 seats
- Minimum term
- 12 months
- Placeholder cost
- £1,450 a month
- Exposure if usage doubles
- Cost holds while headcount is flat
The outcome is conditional. Quote A is cheaper at today's volume and more exposed to growth. Quote B caps the spend but bills for every viewer added. Neither structure wins without a volume forecast the buyer can defend.
Structure fit by buyer profile
- Subscription suits buyers with steady event volumes who need a capped annual cost.
- Seat pricing suits small teams that can name every licensed user and expect flat headcount.
- Usage pricing suits growing sites that can audit the meter month by month.
- Module pricing suits buyers who need one named integration rather than a full suite.
- Managed service suits teams without in-house analysts that can still define acceptance.
- Outcome-linked pricing suits buyers with a defined event and the evidence to dispute it.
The provider shortlist explains why none of these vendors is named best, and what evidence would change that.
Costs that sit outside the meter
Buyer labour is the largest hidden line. Budget for event design, identity rules, data extraction and quality review. Add training, procurement, legal review, incident response and exit.
The NCSC's cloud security principles show how much provider evidence a buyer may need before signing. Where the vendor processes personal data for you, Article 28 terms should cover audit rights, assistance and end-of-contract handling, as the ICO's processor contract guidance sets out.
HMRC's VAT business guidance does not settle the tax treatment of a particular software supply. Keep VAT treatment and recoverability separate, then take advice from a qualified UK tax professional.
Read the service standards notes on precise clocks and blank thresholds before promising a response time to any supplier.
Checks before comparing quotes
- Fix the job and the required output in writing.
- Ask every vendor for the same fieldsedition, unit, included volume, overage, term and exit.
- Run one month of your own data through the proposed meter.
- Add buyer labour, security, privacy and exit costs to each option.
- Mark each option pass, fail or unresolved, and record the reason.
Common questions
How much do attribution platforms actually cost?
Most enterprise suites publish nothing, so the honest answer is quote only. Google Analytics 4 standard is free, and Mixpanel and Amplitude publish free entry tiers with caps you must confirm.
What should I check before comparing usage-based quotes?
Check how the unit is counted, which timezone closes the period, and how retries, duplicates and deleted records are handled. Ask for an export that lets you rebuild the invoice figure yourself.
Which pricing structure is cheapest?
None is cheapest in general. Seat pricing wins when headcount is flat and traffic grows. Usage pricing wins when volume is stable and low. Outcome-linked pricing only looks cheap when the measured event is genuinely incremental.


