Comparing email software when the verdict is a process rather than a winner

Email marketing software reviews are useful only when they preserve the buyer's decision, the evidence available and the gaps nobody checked. A verdict detached from a named plan, workflow and date turns a product feature into a false promise.

This comparison of email software is for an organisation operating from England. Privacy and electronic-marketing sources apply across the UK. Procurement and accessibility duties depend on the buyer and service context, and research closed on 5 September 2026.

No account was opened and no message was sent. No supplier funding, affiliate arrangement or commercial relationship sits behind the draft. Publication remains blocked until named editorial, legal, data-protection, security and accessibility reviewers complete their work.

What to take away

The verdict is a process, not a winner

There is not enough evidence here to name a best platform for England. Two products may show similar campaign features while counting contacts differently, offering different contract terms or leaving different work with the customer.

Brevo prices on monthly send volume, for instance, while Mailchimp prices largely on stored contacts. A high feature count does not offset an unacceptable transfer, inaccessible workflow, weak exit route or unsuitable electronic-marketing process.

Start by describing one real job, naming the sender, recipient population, message purpose, acquisition route and volume pattern. Add content approvals, data fields, integrations, retention need and the people who will operate the service.

The alternative is not always another supplier, but may be a controlled manual process, a smaller existing system or no campaign at all. That baseline in the comparison shows whether software removes work or merely moves it.

The UK's Technology Code of Practice is written for government technology, not as a general commercial accreditation. Its sequence of defining user needs, accessibility, security and privacy is a useful checklist for structuring a buyer's own evidence, and so are its integration, data use and purchasing strategy steps.

Write the review record before reading sales pages

A defensible record identifies:

Desk research can confirm what a supplier currently publishes. It cannot prove that a workflow is usable, that an integration behaves as expected, that support will resolve an incident or that an email reaches an inbox. A sales demonstration is also supplier-controlled evidence. Record it as such.

Treat reviews carefully as the CMA's guidance for businesses and agencies, updated August 2025, covers fake and misleading review practices, but its existence does not authenticate any review.

Keep review platform, collection route, moderation policy and date with each observation. Record the incentive and purchase verification status, and never treat star averages as proof of security, compliance or suitability.

Compare every option on the same questions

Use common criteria and define what would pass each one before seeing the answers. A practical comparison can cover:

Evidence quality matters as much as presence. Put signed contract terms and independently scoped assurance above undated marketing copy. Retain the version because supplier pages change. Where evidence conflicts, show both records and do not choose the more favourable one silently.

A bounded desk comparison of named options

To demonstrate the method, this guide looked at seven named products and a no-new-software baseline. They were included because current first-party pages describe a plan family and an export route. They are not market leaders. Other products were excluded because a complete market search was outside this article's scope.

Brevo's first-party plan description sets out a Standard edition within its current plan family. Its separate contact-export instructions describe CSV export choices.

Mailchimp's plan guide describes its Standard marketing plan and explains that plan type and contact tier affect billing. Mailchimp also documents audience and contact exports.

Five more options appear on most UK shortlists. The table records what each vendor's own pages describe, so verify every entry on the current page before you rely on it.

Entry tiers for a few hundred contacts commonly sit between roughly £8 and £45 a month before VAT as of 2026. Contact tier, monthly send volume, feature level and bundled onboarding drive the climb.

Brevo's send-volume model can cost less for a large list that receives few emails. Contact-based models such as Mailchimp, MailerLite, Campaign Monitor, Klaviyo and HubSpot can cost less for a small list that receives many. These are labelled typical ranges, not quotes.

These are supplier claims and procedures, not independent findings. A buyer in England still needs a dated quote for the proposed configuration, currency and tax treatment.

The manual baseline means keeping the current approved sending process without acquiring another platform. Its cost is not zero: staff time, errors, audit records and existing licences must be counted. It remains a legitimate control option where volumes are small, evidence is incomplete or the proposed automation adds disproportionate risk.

The desk evidence does not support a winner. It establishes questions for a prequalification round and a controlled trial.

Keep legal permission outside the product score

A platform can store an address or trigger a message without establishing permission to use it. The ICO's updated electronic-mail marketing guidance distinguishes subscriber types and addresses consent and the conditions for soft opt-ins. The buyer must classify the actual message, sender, recipient and acquisition route. A supplier's consent flag does not perform that analysis.

Tracking is another decision. Pixels, redirects and related functions may involve storage or access on a device. The ICO's storage and access technology guidance should be applied to each technology and purpose. Turning tracking off may alter reported metrics, and turning it on does not become lawful because a dashboard needs data.

Controller and processor labels must follow the facts. The ICO's contracts guidance describes required subject matter, duration, instructions, confidentiality and security. It covers sub-processing, assistance, deletion or return and audit provisions when a controller uses a processor.

The page is marked under review following the Data (Use and Access) Act, so a qualified UK data-protection lawyer must recheck the current position before publication or signature.

Demand security and accessibility evidence separately

Security assessment should match disclosure, corruption or outage effects. The NCSC's cloud security principles cover data protection in transit, resilience, customer separation and governance. They cover supply chain, identity and audit information.

They are assessment prompts, not certification of any named option here. Ask who reviewed the supplier evidence, what product and region it covers, when it expires and what the customer must configure.

Accessibility needs two test tracks. Operators must use the administration interface. Recipients must understand and act on the output.

The W3C's accessibility standards overview distinguishes guidance for web content, authoring tools and user agents, including WCAG 2.2. Apply a defined standard with an accessibility specialist, representative tasks and assistive technology. A template preview or automated scan alone does not establish an accessible service.

For services to the public, the government's Equality Act guide for service providers explains the anticipatory nature of reasonable adjustments in its scope. Whether and how that duty applies to a particular email journey needs qualified advice. It is not a platform feature comparison.

Test the shortlist without contacting real recipients

Prequalification should remove options that fail non-negotiable evidence gates. Only then run a sandbox exercise with synthetic contacts and content. Ask the supplier to identify the exact edition and region used.

Test role separation, approval, authentication setup, duplicate import handling and suppression propagation. Then test tracking controls, accessible content production, audit export and incident escalation. Finish with complete export, deletion and rollback.

Record expected state, observed state, evidence location, operator and timestamp, and note the defect owner. Do not use live addresses or infer deliverability from a test mailbox. A passed scenario shows only that the tested configuration performed that task at that time.

Set exit conditions before the exercise. Stop if a suppression state is lost, privileged activity is not attributable, required records cannot be exported, a critical operator journey is inaccessible or the supplier will not resolve a contractual unknown. Recovery must restore the prior state and reconcile every affected record.

Avoid false precision in scoring

A total score can conceal a fatal defect. Use pass, fail, unknown and not applicable for non-compensating gates. Weighted scores may help among options that passed, but publish the weights, evidence rule and sensitivity. If changing a plausible weight reverses the order, report an unstable preference rather than a winner.

Claims such as "best", "most secure" or "easiest" also create an evidence problem. CAP's substantiation guidance explains that objective claims capable of substantiation require documentary evidence before publication. Editorial comparison and promotional reuse should be reviewed separately. A transparent method cannot rescue a conclusion that the underlying evidence does not support.

The decision pack

The final pack should contain the workflow, option register, evidence index, unknowns and gate decisions. Add trial results, contract deviations, whole-life cost model and migration plan. Add the exit rehearsal and signed approval.

Record the exact contracting entity from the proposed agreement. The Companies House information service can provide company filings and officer details, but those records do not prove product capability, solvency, authority to sell a brand or contract suitability.

A sensible decision can be "not yet", so hold the purchase if the reviewer is unassigned, critical evidence is missing, or the buyer cannot recover from failure.

Re-run the review when the edition, price basis, data route, sub-processor list or terms change, and repeat when law, workflow or business population changes.

The result is an auditable selection, not permanent endorsement.

Before you act

Common questions

Why does the guide refuse to name a best email platform?

There is not enough evidence to name a best platform for organisations in England. Two products may count contacts differently, offer different contract terms or leave different work with the customer. A high feature count cannot offset an unacceptable transfer, inaccessible workflow, weak exit route or unsuitable electronic-marketing process.

Which email platforms does this guide compare?

Seven named products appear in the bounded desk comparison: Brevo, Mailchimp, MailerLite, Klaviyo, Campaign Monitor, Dotdigital and HubSpot Marketing Hub. The manual baseline sits alongside them. Each is recorded with its published pricing basis, who it suits and its documented export route, not a score.

What can desk research actually confirm?

Desk research can confirm what a supplier currently publishes. It cannot prove a workflow is usable, an integration behaves as expected, support will resolve an incident or an email reaches an inbox. A sales demonstration is also supplier-controlled evidence, so record it as such.

Why is the manual baseline included in the comparison?

The baseline keeps the current approved sending process without acquiring another platform. Its cost is not zero: staff time, errors, audit records and existing licences must be counted. It remains a legitimate option where volumes are small, evidence is incomplete or automation adds disproportionate risk.