Skip to content
All systems operationalv0.16.68FREE TIER · 3 DEVICESTENVO AI · MCP FOR CLAUDE & CURSOR
Back to BlogOpinion

Vendor Lock-In: Price Your Exit Before You Sign

Tenvo Editorial Team9 min read
Vendor Lock-In: Price Your Exit Before You Sign

Nobody signs a remote access contract planning to leave. That is precisely why the exit is never priced. Three years later the renewal quote arrives with a number nobody expected, somebody says "fine, we'll move", and the team spends the next fortnight discovering what moving actually costs.

Nobody signs a remote access contract planning to leave. That is precisely why the exit is never priced. Three years later the renewal quote arrives with a number nobody expected, somebody says "fine, we'll move", and the team spends the next fortnight discovering what moving actually costs.

Lock-in is rarely a clause. There is no paragraph that says you may not leave. What there is instead is a pile of small, boring facts: the device IDs are in the old vendor's database, the technicians' habits are built around the old console, the audit trail your auditor asked for lives in a product you are about to stop paying for. Each fact is cheap on its own. Together they are the switching cost, and the vendor does not have to engineer them deliberately for them to work in the vendor's favour.

The useful move is to price the exit while you still have leverage — that is, before the signature, when the salesperson still wants something from you.

The five things that actually hold you

When teams say they are locked in, they almost always mean one of these. It is worth naming them separately, because they have very different price tags.

1. The device list

Unattended access is the expensive one. Attended support — a one-off session where somebody on the far end reads you a code — costs nothing to switch. Unattended access means an agent is installed and enrolled on every machine you support, usually with an identity the vendor issued. Moving means touching every endpoint again.

For 40 office machines on one LAN that is an afternoon. For 900 machines across clients' sites, half of them laptops that appear on the network unpredictably, it is a project with a budget. Ask yourself the honest question before signing: if we had to re-enrol every device, who does it, and over how many weeks? If the answer is "we genuinely cannot", you are not buying a tool, you are buying a dependency.

2. The credential store

If your technicians have been saving target credentials inside the vendor's console, those credentials are now in a vault you are leaving. Some products export them. Some export them only as a list of hostnames with the secrets stripped — which is the same as not exporting them. Then somebody spends a week asking clients for passwords they do not remember setting.

This is the argument for keeping secrets somewhere that is not the access tool in the first place. We go into the trade-offs in credential vaulting for MSPs, but the short version is that a vault you own survives a vendor change and a vault you rent does not.

3. The audit history

This is the one that surprises people. Session logs — who connected to what, when, for how long, who approved it — are usually retained by the vendor and shown to you through their interface. Cancel the subscription and the interface goes away.

If you are in scope for anything with a retention requirement, that history is not a nice-to-have. ISO 27001, SOC 2 and PCI DSS all expect you to be able to produce access records for a defined period, and "our previous supplier had them" is not an answer an assessor accepts. See remote desktop audit logging for what a usable record contains, and ISO 27001 remote access controls for the retention side.

The fix is unglamorous: export the log on a schedule to storage you control, from day one, whether or not you ever plan to leave. A nightly export costs an hour to set up and removes an entire category of exit cost.

4. The integrations

Every connection you build into the access tool is a small weld. The ticketing system that opens a session from a ticket. The SSO that provisions technicians. The RMM that triggers a connection from an alert. The script somebody wrote in 2024 that nobody has looked at since.

Count them before you sign, and again every year. Teams are routinely wrong about this number by a factor of two, because the welds get added one at a time by whoever needed them.

5. Habit

The least measurable and often the most expensive. A support team that has used one console for four years is fast in it. Move them and first-call resolution drops for a month while everybody relearns where the file transfer button is. That is a real cost, it shows up in your support metrics, and it is the reason some teams stay on a tool they have otherwise outgrown.

You cannot eliminate it. You can stop it from being a surprise by treating retraining as a line in the migration budget rather than a thing that will somehow just happen.

What to ask before you sign

These are the questions that change the shape of the deal rather than the price on the front page. Ask them in writing, and keep the answers.

AskWhat a usable answer sounds like
How do we export the device list?A named format — CSV or an API endpoint — available to us at any time, including after cancellation notice.
How do we export session logs, and for how long are they available?A documented export with a stated retention period, not "contact support and we'll see".
What happens to data at the end of the contract?A window, in days, during which we can still retrieve it, stated in the agreement.
Can we reduce seats or devices mid-term?Yes, with notice, down to some stated floor. "Only at renewal" means you pay for your own shrinkage.
What is the notice period, and does it auto-renew?Thirty days against a twelve-month term is normal. Ninety days against a three-year auto-renewal is a trap you walk into once.
What caps price increases at renewal?A number or an index. "We've never raised prices much" is not a cap.
What triggers a licence audit, and what does one cost us?Clear usage thresholds. Vague "commercial use" detection is a bill waiting to happen — see how commercial-use detection works.

None of this is adversarial. A vendor confident in the product answers all seven without flinching, because they expect you to stay for reasons other than difficulty of leaving. Hesitation on the export questions in particular tells you something worth knowing.

The notice period deserves its own paragraph

Of everything on that list, the auto-renewal plus notice-period combination causes the most avoidable pain. A ninety-day notice window on an annual auto-renewal means your real decision date is nine months after you sign, not twelve — and it is a date nobody has in a calendar. Miss it and you have bought another year of a tool you decided against.

Put the notice deadline in a shared calendar on the day you sign, with a reminder a month before it. This takes two minutes and is the highest-return thing in this article.

"If they squeeze us, we'll just self-host"

This is the most common exit plan we hear, and it is usually not an exit plan. Standing up your own relay infrastructure does not remove a bill, it swaps a predictable one for a set of obligations: somebody patches the server, somebody holds the keys, somebody is awake when it falls over at 2am, and somebody explains to a client why a single-region deployment had no failover. We priced that out in the hidden costs of self-hosting, and in total cost of ownership.

Self-hosting earns its place when a written requirement forces it — data residency in a specific jurisdiction, an air-gapped network, a contract clause you cannot negotiate away. Those are real and they are not rare. What it is not is the neutral fallback, and treating it as the free option is how teams end up running infrastructure they never budgeted for, staffed by one person who then goes on holiday.

If your leverage in a negotiation rests on a credible threat to self-host, make sure the threat is actually credible: name the person who would operate it and the hours it would take them per month. If you cannot, the vendor can tell, and so should you.

What low exit cost looks like in practice

Some of this is about the product you choose, and some of it is about how you use whatever you choose. The second part is entirely in your hands:

  • Secrets live outside the access tool. A vault you control, with the access tool reading from it rather than holding it.
  • Logs are exported on a schedule to your own storage. Nightly, automated, verified occasionally by actually reading one.
  • The device inventory exists somewhere else too. Your asset register is the source of truth; the vendor's list is a copy.
  • Integrations are documented. A one-page list of every weld, updated when somebody adds one.
  • Deployment is scripted, not manual. If you can push an agent to 900 machines with a script, re-enrolment is a scheduling problem rather than a staffing one. See deploying by MSI and unattended access at fleet scale.

A team that has done those five things can change supplier in a fortnight. A team that has done none of them is quoting a six-week project to leave, which means in practice they will not leave, which means the renewal price is whatever the vendor decides.

On the product side

The signals worth looking for are mostly about whether the vendor has pre-committed to things that are inconvenient for them. Published pricing you can read without a sales call. A documented export path. An open-source client, which at minimum means the thing running on your endpoints can be inspected and is not a black box you lose access to. A licence you can read — if it is open source, the AGPL guide covers what that obliges of you in return.

Tenvo is built that way on purpose: the client is open source under AGPL-3.0, pricing is on the pricing page rather than behind a form, and device and log export are documented rather than discretionary. We would rather compete on whether the product is good than on how annoying it is to leave. That is also, bluntly, the self-interested position — a vendor that keeps customers by friction stops improving.

A short checklist before signing

  • Get written answers to the seven questions above. Keep them with the contract.
  • Put the notice deadline in a calendar the day you sign.
  • Confirm the data-retrieval window after cancellation, in days.
  • Confirm whether you can reduce seats mid-term, and to what floor.
  • Confirm what caps renewal increases.
  • Decide where secrets and logs will live — outside the tool — and set that up in week one, not year three.
  • Write down, in one line, who would re-enrol the fleet and how long it would take. If you cannot, fix that before it becomes the answer to a negotiation.

An hour of this before signing is worth more than any amount of regret at renewal. If you are earlier in the process, how to buy remote desktop software covers the evaluation itself, the procurement questions go deeper on pulling vendors apart, and a staged migration plan is what the exit actually looks like when you do have to run it.

Get Tenvo

Ready to try it yourself?

Free for 3 devices, no credit card. Up and connected in two minutes.

All articles