marketing, strategy, marketingstrategy, business, flatlay, conversion, pricing, ironing, automation, marketingautomation, wood, brown marketing, marketing, conversion, pricing, pricing, pricing, pricing, pricing
Photo by Campaign_Creators on Pixabay

Costs and pricing

Part of A measurement budget method that keeps three separate ledgers and tests its benefits

Comparing measurement pricing structures by following the incentive

Compare measurement pricing structures on one whole-life basis, including usage, staffing, risk allocation, migration, support, tax and exit.

What to take away

  • Compare structures on whole-life cost, not the headline charge: add buyer labour, privacy review, security checks, migration and exit.
  • Follow the incentive. Usage charges reward more billable volume, fixed projects reward a tight scope, time-based work rewards more hours, subscriptions reward unchanged limits.
  • Record the switching value, the event count or consultant hours at which a rival structure becomes cheaper, before you sign.
  • Cost alone cannot waive a failed privacy, security or delivery gate.

Five pricing structures and their billable units

Five shapes dominate privacy-first measurement contracts. The pricing models compared explains how each one moves cost between buyer and seller. No tender, rate card or product was reviewed there, so treat named suppliers as contract shapes rather than endorsements.

[figure 1]

Five measurement pricing structures

Structure

Flat subscription
£ per fixed period
Usage charge
£ per event or unit
Seat charge
£ per user
Fixed project
£ per project
Time-based
£ per hour

Billable unit

Flat subscription
Which limits change access?
Usage charge
How is use disputed?
Seat charge
Who counts as a user?
Fixed project
Is scope precise?
Time-based
Who manages priorities?

Main buyer question

Flat subscription
Usage charge
Seat charge
Fixed project
Time-based

Five pricing structures

Billable unit

Flat subscription
£ per fixed period
Usage charge
£ per event, request or data unit
Seat charge
£ per authorised user per period
Fixed project
£ per accepted deliverable
Time-based service
£ per hour or day

Named examples

Flat subscription
Plausible Analytics, Fathom Analytics, Simple Analytics
Usage charge
PostHog, Piwik PRO, Snowplow
Seat charge
Adobe Analytics enterprise
Fixed project
Measurelab, InfoTrust, Bounteous
Time-based service
Upwork, Toptal, Measurelab

Buyer question

Flat subscription
Which limit or fair-use term can change access?
Usage charge
How is billable use measured, and how is it disputed?
Seat charge
Are viewers, service accounts and contractors counted?
Fixed project
Are inputs, changes and acceptance defined?
Time-based service
Who approves effort, and which role performs it?

The Cabinet Office risk allocation and pricing guidance is written for government contracting. It shows how pricing design shifts risk, not a rule for private analytics agreements.

What each structure rewards

A usage model fits variable demand but rewards more billable volume. A subscription stays predictable until a limit, renewal or necessary add-on changes the basis. Each shape leaves the buyer exposed in a different place.

Ask which party can observe the billable meter and how the reading can be challenged. Put overage, credit expiry, pause rights, unused capacity and supplier-caused rework through the same test. The contract questions for a measurement supplier cover exit, data roles and acceptance terms.

Fixed projects move delivery risk only when scope and acceptance are precise. Separate a legitimate change in scope from correction of a defective deliverable. Time-based work puts effort on the invoice and leaves priority-setting with the buyer, so require role, task, hours and approval on each record.

[figure 3]

Where each structure puts risk

Structure

Usage charge
More billable volume
Fixed project
Delivery risk shift
Time-based
Billable effort
Subscription
Predictability

Rewards

Usage charge
Meter disputes
Fixed project
Scope precision
Time-based
Priority management
Subscription
Limits and add-ons

Buyer exposure

Usage charge
Fixed project
Time-based
Subscription

Example: a 100,000-pageview site priced four ways

Figures below are arithmetic from published list prices, not a quote.

100,000-pageview site priced four ways

Billable unit

Flat subscription
£ per month per site
Usage charge
£ per event
Fixed project
£ per accepted deliverable
Time-based
$ per hour

Modelled 12-month cost

Flat subscription
about £144
Usage charge
£0 up to a paid-tier figure the buyer must model
Fixed project
£3,000 to £6,000 once
Time-based
$1,200 to $6,000 for 40 hours

Basis

Flat subscription
Fathom Analytics plan at about £12 a month for 100,000 monthly pageviews
Usage charge
PostHog free to 1 million events a month
Fixed project
five days at a UK rate of £600 to £1,200 a day
Time-based
freelance rates of $30 to $150 an hour

The gap between £144 and £6,000 is the point. Subscription, usage and project pricing answer different questions, and only the subscription figure carries no implementation labour.

Normalising proposals before scoring

  1. Convert every proposal to one currency, contract term and VAT basis.
  2. Convert units. A Plausible pageview, a PostHog event and an Adobe server call are not the same measure.
  3. Add minimum commitments and buyer-requested options, since both bind the buyer.
  4. Exclude internal labour and recoverable VAT here, then add both for separate cash and economic views.
  5. Test lower and higher usage, delayed launch, added seats, change requests and early termination.

HMRC's VAT charging page gives standard-rate context, but a finance or tax reviewer decides the treatment. Never compare a gross quote with a net one.

The measurement budget template keeps cash, economic cost and benefit on separate sheets, which is what a whole-life comparison needs.

Vendor list prices and their billable units

  • Plausible Analyticsabout £7 to £9 a month per site for 10,000 monthly pageviews.
  • Fathom Analyticsabout $15 a month (about £12) per site for 100,000 monthly pageviews.
  • Simple Analyticsabout $19 a month per site for 100,000 pageviews.
  • Matomo Cloudabout $23 a month for 50,000 hits.
  • Piwik PROfree to 500,000 events a month, then paid plans from about $99 a month.
  • PostHogfree to 1 million events a month, then usage-based pricing.
  • Snowplowpriced per event volume, with a self-hosted open-source option.
  • Google Analytics 4: free up to 10 million hits a month. GA360 enterprise contracts typically start in the tens of thousands of pounds a year.
  • Adobe Analyticsseat and server-call model, typically £30,000 to £100,000+ a year.
  • Fixed project workMeasurelab, InfoTrust and Bounteous, with UK analytics implementation day rates typically £600 to £1,200.
  • Time-based workUpwork or Toptal, $30 to $150 an hour.

Matomo and Plausible allow data export; PostHog and Snowplow allow self-hosting or export. Check notice periods and deletion policies.

Public bodies eligible for G-Cloud can use the G-Cloud buyers' guide. Private buyers should not imply eligibility, and a framework listing does not make a service suitable.

Common questions

What should be normalised before comparing measurement pricing proposals?

Convert each proposal to the same currency, contract term and VAT basis, then use expected units the buyer can evidence. Include minimum commitments and buyer-requested options, and keep internal labour and recoverable VAT in separate views.

How does a usage charge create a different incentive from a flat subscription?

Usage pricing fits variable demand but rewards more billable volume, so the supplier benefits when events rise. A subscription is predictable until a limit, renewal or necessary add-on changes the basis.

What costs sit outside the headline price?

Buyer labour, privacy review, security checks, migration and exit. Add internal labour and recoverable VAT separately for cash and economic views.

When should the comparison be reopened?

If a minimum commitment, indexation clause, unit definition, service scope or exit charge changes. A lower modelled bill is not evidence of better measurement.

More in Costs and pricing