The licence is the part everybody compares and rarely the part that decides the total.

What MSP software actually costs in the UK

The pricing models used in this market, the costs that are usually left out of the comparison, and how to build a total that means something.

Pricing in this market is difficult to compare on purpose in some cases and by accident in others. Products are licensed per technician, per endpoint, per tier or as part of a portfolio arrangement, and the implementation is sometimes included, sometimes quoted separately and sometimes delivered by a third party you have not met yet.

This guide sets out the shapes rather than the numbers. We deliberately do not quote other suppliers’ prices - they change without notice, they vary by territory and reseller, and a stale figure on our site would be both inaccurate and unfair.

The four pricing shapes you will meet

  • Per technician, per month. The most common shape for PSA-type products. Predictable, and it scales with your team rather than with your clients.
  • Per endpoint or per device. More common where monitoring is bundled. Scales with your clients’ estates, which can move faster than your headcount.
  • Tiered by capability. A per-user price that varies by which functional tier you need. The risk is discovering in month nine that a feature you assumed was included sits one tier up.
  • Quoted through a partner or portfolio. Common with the larger platforms. Perfectly legitimate, and it makes like-for-like comparison hard unless you insist on a full written total.

The costs usually left out of the comparison

  • Implementation and configuration, particularly where the platform is highly configurable and the configuration is therefore the project.
  • Data migration, and specifically ticket history - which is often priced separately and which you should not skip.
  • Integration or connector licensing, where each connection to another system carries its own annual charge.
  • Client portal seats, charged per client user. This is the charge that quietly stops providers inviting clients in.
  • Training, both initial and for people who join afterwards.
  • Your own internal time. Usually the largest single cost of any implementation and almost never in anybody’s spreadsheet.
  • The tier upgrade you will need in year two, which should be modelled now rather than discovered later.

Ask every supplier for a single document showing the full three-year total including all of the above. The ones who provide it readily are telling you something, and so are the ones who do not.

How to think about implementation cost

Implementation effort scales with two things: how much configuration the platform requires before it is usable, and how unusual your operation is. A highly configurable product is more capable and more expensive to set up, and both halves of that are true at once.

What matters most is the shape of the arrangement. A fixed price per stage, with what each stage delivers written down, is a supplier taking the estimation risk. A day rate with an estimate is you taking it. Both are legitimate, but they are very different propositions and the second one has no natural end.

What we charge, plainly

CubeMSP is priced per named technician per month, with one edition and no functional tiers. Client users invited to shared boards or reviews are not charged for. The API is included and is not licensed per connection.

Implementation is a one-off fixed fee covering discovery, configuration, data migration and training, quoted after discovery so that it cannot quietly become an open-ended project. Ticket history migration is part of it rather than an extra, because the knowledge features depend on it and charging separately for that would be perverse.

We are in early access, and pricing for early customers reflects both that and the expectation that they will influence the roadmap. Ask us for current figures - they are on the pricing page and we will put them in writing.

Working out whether it pays for itself

Build this by measuring one thing rather than modelling everything. Pick the largest of: renewals missed or auto-rolled in the last year, vendor price rises not passed on, licence counts billed incorrectly in either direction, chargeable work never invoiced, or engineering time spent rediscovering known fixes.

Most providers find that a single one of those, measured properly for one year, exceeds the annual cost of the platform. If none of them does, that is a useful finding and it means you should wait.

In short

What to take away

  • Four pricing shapes: per technician, per endpoint, tiered, or quoted through a partner.

  • Implementation, ticket history migration, connector licensing and portal seats are the usual omissions.

  • Fixed-price staged implementation moves estimation risk to the supplier; a day rate leaves it with you.

  • Measure one specific loss properly rather than modelling a full return on investment.

Questions

Asked while people are working this out

  • Because they change without notice and vary by reseller and territory, so anything we published would be wrong within months. A stale figure on our site is both an accuracy problem and an unfair one.

  • It depends on which grows faster for you. Per-endpoint pricing scales with your clients’ estates, which for a growing provider can move considerably faster than headcount. Model both against your own three-year plan.

  • In our view yes, and staged, with each stage’s deliverable written down. It is the only structure where the supplier carries the estimation risk they are better placed to carry.

  • It is a common model and it has a predictable effect: providers ration invitations, so the portal stays empty and the transparency never materialises. We do not charge for client users, which is a deliberate decision rather than an oversight.

After reading

Bring the questions this raised

If the guide has made you think of something specific about your own business, that is exactly the conversation worth having. No obligation and no scripted demonstration.