How we compare

The problem with software comparisons is that the reader cannot see where the numbers came from. So we publish the whole method — including what the site does not have.

What the comparison rests on

Every figure on the site is of the same kind: read from the provider’s own public source and dated. Prices and plan limits from the price list, where the data is stored and which processors are used from the privacy policy, the billing calculation rules and whether a feature exists from the provider’s own help and documentation pages. The check date sits next to each figure.

That is a checkable fact, not an opinion: you can open the same price list and see the same number. If the number changed after the check date it is out of date, not wrong — and we correct it in the next round.

How the prices are read

None of the price lists in the comparison is a static page: the price comes from a slider, a dropdown or a tab, and many pages show a different currency depending on the country. So the prices are read in a real browser, the same way you would see them, not with a scraper.

For each tool we record three figures — the monthly price at 500, 2,500 and 10,000 contacts — on the cheapest paid plan that supports that list size, excluding VAT. We also record five things that would otherwise make the comparison unfair:

• Which plan the price is for. The cheapest plan does not always include automation, in which case the real comparison price is the price of a more expensive plan.
• Which billing period. Some providers only show the monthly price on annual billing, which is not comparable with monthly billing.
• The smallest list sold. Some do not sell a 500-contact list at all, so even a small list costs the price of a thousand contacts.
• Promotional price or list price. A first-year discount says nothing about what the service costs in year two.
• How contacts are counted. Does an unsubscribed contact still count towards the limit, is the same address counted several times across lists, and what does exceeding the limit cost. That rule is on the provider’s help page rather than the price list, and it changes the cost of the same list more than some of the differences between price tiers do.

These appear as caveats under every price table, not as hidden footnotes.

When the figures are checked

Prices are checked every 90 days and the check date is shown on the page. The interval is a rule rather than a recommendation, because price lists go stale faster than articles do. The first round on 3 September 2026 showed it immediately: MailerLite’s free plan had shrunk to half its former size (500 → 250 subscribers) and GetResponse had removed its free plan altogether, while both are still promised online with the outdated figures. Both changes are recorded, with dates, in the tool’s own changelog, so you can see what changed and when.

What we do not claim

We do not give scores. A numeric score would require every tool to have been used with real data and measured the same way. We have not done that, so there is no score. We also do not call a tool “the easiest”, “the clearest” or “the fastest” — those are measurement results, not opinions, and we have no measurement.

We do not measure deliverability ourselves. A reliable measurement would need dozens of mailboxes across providers and months of tracking. The deliverability figures we publish are measured by a third party: they come from Tooltester’s series of 12 test rounds, and the source is stated under every table. The series ended in January 2024, and in August 2026 Tooltester explained why it will not continue: seed lists proved unreliable and providers started optimising for them. That is why we present the figures as history, not as a measurement of the present, and we say so under every table. Our own contribution is the arithmetic — the average and the range across the whole series, because one round tells you less than the spread does. For two providers (Systeme.io, EmailOctopus) there is no independent measurement at all, and we say so directly where they appear.

We do not assess enterprise price lists that are only given in a sales call, and we do not promise discount codes or promotional prices: they change weekly, and an article built on an offer is out of date before it is read.

These absences are deliberate. A comparison that says what it does not know is more useful than a comparison that invents the missing part.

Independence

Some of our links are ad links, which earn us a commission. The commission does not affect the ordering, and the comparisons also include tools that pay us nothing — you can tell them apart because their links carry no “(ad)” mark. If a tool pays us nothing but is the best choice for the need described, it stays the recommendation. >Read more about how the site is funded →