The comparison, side by side
We sell one of these two options, so treat the table as a starting point and check the salary figures against your own market. The columns are written to be usable by somebody who decides against us.
| Hiring in-house | Using an agency | |
|---|---|---|
| First year cost | Salary plus 25 to 40 per cent, plus recruitment and a two to four month vacancy | The project fee only, with no employment commitment |
| Time to the first system running | Three to six months, most of it hiring and onboarding | Two to eight weeks, depending on scope |
| Breadth of skill | One person's range, which rarely covers integration, voice, extraction and security equally | A team that has already solved the common problems elsewhere |
| Knowledge of your business | Deep, and compounding every month | Shallow at the start. Adequate by the end of discovery if the discovery is done properly |
| Continuity risk | High. One resignation stops all maintenance | Moderate. A supplier can also disappear, which is why ownership matters |
| Who owns the result | You, unambiguously | You, only if it is built on your accounts and documented. Confirm in writing |
| Cost of the eleventh automation | Marginal. The salary is already committed | Another project fee, unless a retainer changes the shape |
| Best fit | Continuous, permanent automation demand | A defined set of processes, or a first build before the demand is proven |
The numbers, at typical mid-2026 rates
Approximate and market dependent. Advertised United Kingdom ranges for automation and integration engineers in mid-2026 sit broadly between GBP 45,000 and GBP 75,000, with senior specialists between GBP 70,000 and GBP 110,000. Gulf packages vary widely and are frequently structured differently, so a like for like comparison has to be done locally.
| Year one | In-house automation engineer | Agency, defined scope |
|---|---|---|
| Direct cost | GBP 60,000 salary, call it GBP 78,000 loaded | One to three projects, commonly USD 2,500 to 15,000 each |
| Recruitment | 15 to 25 per cent of salary through an agency, or several months of internal effort | None |
| Idle time | Paid whether or not there is a queue of work | None. You buy what you scope |
| Output in year one | Typically three to six systems once ramped up | As many as you commission and pay for |
| What you hold at the end | A person with deep context, and everything they built | The systems, the documentation and the accounts, if you insisted on them |
When hiring is plainly the better answer
Said without hedging, because these cases are common and an agency is the wrong recommendation in every one of them.
You have more than about ten processes worth automating
At that volume the salary is cheaper than the project fees, and it stops being close. The break-even is usually somewhere between the sixth and the twelfth system.
Your systems change constantly
If your internal tools are under active development, automation needs continuous adjustment. Paying a supplier per change is worse value and slower than having somebody inside who sees the change coming.
The domain knowledge takes months to acquire
Complex regulated operations, unusual pricing rules and specialised workflows all favour somebody who lives inside the business. An external supplier can learn it, but you pay for that learning every time you engage a new one.
Automation is becoming part of your product
If what you are building will be sold to your own customers rather than used internally, it is engineering, not a service purchase, and it should be owned by an employed team.
You already have engineers with spare capacity
A capable developer can learn this quickly in 2026. If you have one who is interested and available, a training budget and a few weeks is usually a better investment than a first project fee.
When an agency is the better answer
- You want the first system running in weeks and cannot wait out a hiring cycle.
- You have two or three processes to automate and no evidence yet of a continuing pipeline. Hiring for demand you have not proved is the more expensive mistake.
- The work spans disciplines one person is unlikely to cover well, particularly voice, document extraction and security review at the same time.
- You need a fixed cost you can approve, rather than an open commitment on the payroll.
- Nobody currently inside the business could interview a candidate for the role, which is a common and under-acknowledged reason first hires in a new discipline go badly.
The option most organisations actually end up at
Sequential rather than either or. The first two or three systems are delivered externally, on your accounts, with documentation and a recorded handover. That proves whether the demand is real, produces working examples, and establishes the patterns. When the pipeline justifies a salary, the hire arrives to a working estate rather than a blank page, and their first three months are productive instead of exploratory.
The reverse order, hiring first and commissioning externally later, usually goes worse. The new hire spends the first quarter deciding on an approach, and any external supplier brought in afterwards has to work around decisions taken before there was any evidence.
What to insist on either way
- Every system built on accounts in your organisation's name, with your billing details.
- Written documentation of what each system does, what it connects to and what happens when it fails.
- A recorded handover session, which costs an hour and saves weeks when the person who built it is unavailable.
- A named owner for each automation inside your business, even when an external team maintains it.
- Monitoring that alerts somebody employed by you, not only the supplier.
Our own position, marked as ours
We are an agency, so the balance of this page is deliberately weighted against our own interest in places. Our honest view is that organisations with continuous automation demand should hire, and that a first build with an external team is the cheaper way to find out whether that demand exists. Everything we build sits on accounts in your name and is documented so that it can be maintained by somebody else, including a person you hire later. Our fees are fixed and published, and the support plan is optional and never a condition of the build.
Frequently asked questions
What is the real first-year cost of hiring?
Base salary plus employer contributions, equipment, tooling and recruitment, which typically adds 25 to 40 per cent on top of the advertised figure. Add the vacancy period: a specialist automation hire commonly takes two to four months to fill, during which nothing is being built.
What is the biggest risk of using an agency?
Dependency. If the systems are built inside a supplier's platform, or documented only in their heads, you cannot leave without losing the work. The way to remove that risk is contractual and practical: everything built on accounts in your name, documentation handed over, and a working handover session recorded.
Can one person cover the whole discipline?
Rarely at first. A production automation practice touches integration, voice, data extraction, security and monitoring, and few individuals are strong across all of them. This is the honest argument for an agency at the start and for a hire once the patterns are established and repeatable.
What happens if the person we hire leaves?
Whatever they built stops being maintained the day they go. A single in-house specialist is a concentration risk that most organisations only notice after the resignation. Written documentation and a second person who can read the systems are the mitigations, and both are usually skipped.
Is there a sensible middle path?
The most common successful pattern is to have the first builds delivered externally, insist on ownership and documentation, and hire once there is enough running to justify a salary. The new hire then inherits working systems and a pattern to follow instead of a blank page.
Related answers
Disclosure, and the only sales pitch on this page
This page is published by a company that sells the thing it describes.
Oxford Crown Technologies builds voice and operations agents for organisations headquartered in London and across the Gulf. We publish these pages as reference material, including market figures that are not ours and that do not always favour us, because a buyer who understands the range negotiates better with everybody, including us. Leadership holding postgraduate degrees from the University of Manchester and the University of Oxford.
Our own builds are fixed fee, live in 14 days, built on accounts in your name, and carry a 28 day money back guarantee. We prove Return on Investment, or you do not pay.
Reviewed 25 July 2026. Market figures change; where this page quotes a range from a third party, the source is named above. Figures described as ours are our published fees or typical results for a first build, not measured results for a named client.