A supplier once put an analyser in front of a clinic manager I was advising and slid a single number across the table: zero. No capital cost, no purchase order to defend, a machine that could produce results in the room from that afternoon. She was delighted. The finance director was delighted. The device was installed inside a fortnight, and for about three months everyone believed they had won.
Then the reagent invoices started arriving. Then the controls. Then a service visit that was not quite covered by the plan they thought they had signed. Then the day a cartridge lot expired in the fridge because throughput had been lower than the vendor's forecast, and a box of them went in the clinical waste unused. Eighteen months in, the "free" analyser was quietly one of the more expensive lines in the department, and switching away from it meant unpicking a contract, retraining staff and rebuilding the data feed. It was free the way a printer is free.
This is the central claim of this piece, and I will defend it hard: the capital price of a point-of-care analyser is the least important number in the whole deal. Procurement that optimises the sticker price is optimising the wrong variable. The money lives downstream, in reagents, controls, calibrators, external quality assessment, service, connectivity, wastage and staff time, and it recovers the vendor's margin many times over across the life of the contract. Buy on total cost of ownership and cost per reportable result, or you will buy badly and feel clever doing it.
Why "free" is a pricing strategy, not a gift
Reagent rental and managed-service contracts are completely standard commercial models in in-vitro diagnostics, and there is nothing sinister about them. A vendor places the analyser at low or no capital cost, and recovers its investment through the consumables and service you buy over the contract term. It is the same logic as a mobile phone bundled into a two-year airtime plan, or a coffee machine given away against the pods. The hardware is the hook; the annuity is the business.
The reason this matters for a buyer is not that the model is unfair. It is that the model deliberately moves the cost out of the one number your finance process scrutinises most, the capital price, and into a stream of smaller recurring numbers that no single approval ever examines in aggregate. A capital purchase gets a business case, a committee and a challenge. A monthly reagent order gets a signature. The economics of the device are decided in the place your governance is weakest, and they are decided by default, months after the machine arrived, when nobody remembers it as a purchasing decision at all.
I have watched capable finance teams win a hard fight over three thousand pounds of capital and then wave through, unexamined, a consumables commitment worth many times that over the contract. The rigour was real. It was simply pointed at the wrong number. That is the trap in one sentence: the scrutiny is fierce, and it lands nowhere near the money.
The hardware is the hook. The annuity is the business. And the annuity is priced where nobody is looking.
So the discipline is simple to state and hard to practise: never let the capital price anchor the decision. Treat it as one line among many, and often a small one.
The lines below the waterline
Once you stop looking at the sticker, the real cost structure comes into view. Most of it is invisible on day one, which is precisely why it survives the buying decision.
Reagents and consumables
This is the main engine of the annuity. Every reportable result consumes a cartridge, a strip or a reagent pack, and the price per test is where the vendor makes its return. A low headline price is very often mirrored by a high consumable price, because the two are set together. The number that actually governs your budget is cost per test multiplied by realistic annual volume, not the machine's face value.
Controls, calibrators and EQA
Quality is not free and should never be treated as optional. Internal quality control materials are consumed every day the device runs, calibrators are consumed on a schedule the method dictates, and external quality assessment scheme fees, through UK NEQAS, WEQAS or equivalent, recur every year for every analyte in scope. These are the costs of being able to trust a result, and they scale with the number of analytes and the QC frequency your risk assessment demands. A cheap device that measures more analytes can carry more quality cost, not less.
Cartridge wastage and short shelf life
Unit-use consumables have expiry dates, and reagent packs, once opened, have on-board stability limits measured in days or weeks. A device sized for a busy clinic, placed in a quiet one, throws away reagent it never used. Wastage is a real cost that vendor forecasts rarely dwell on, because the vendor is paid whether the cartridge produces a result or a bin liner. Low, lumpy or seasonal throughput turns a good per-test price into a bad one.
Service, maintenance and downtime
Maintenance and service contracts are a recurring line in their own right, and the coverage varies enormously between an all-inclusive plan and a bare warranty that bills for parts and visits. Downtime has its own cost that never appears on any invoice: when the only analyser is broken, the work goes to the main laboratory, to a courier, or to a send-away with a turnaround that changes the clinical decision. A cheaper device with worse uptime can cost more in disruption than it ever saved in capital.
Connectivity and data
A result that a clinician cannot see in the record safely is only half a result. Getting numbers off the device and into the patient record involves interfacing effort, sometimes licence fees, and the standing cost of keeping that feed alive through software updates on both sides. Some low-capital placements assume manual transcription, which is the most expensive and least safe option of all once you count the staff time and the transcription errors it invites.
Staff time
The most consistently ignored line is people. Every QC run, every calibration, every maintenance task, every competency assessment and every EQA return is someone's paid time. It does not show on a purchase order and it is rarely modelled, yet across a fleet and a five-year term it is one of the largest costs in the whole exercise. A device that demands more manual QC or fiddlier maintenance is quietly spending your establishment.
The crossover: why cheap on day one is dear by year five
Put these lines on a timeline and the picture inverts. Consider, for the sake of argument, two analysers. Device A is placed free but carries a high per-test consumable price and modest QC and service costs. Device B costs a real capital sum up front, illustrative say a few thousand pounds, but its consumables and service are cheaper. On day one, A is obviously the winner. It stays the winner for a while. Then the cumulative cost curves cross.
The exact crossover date depends entirely on volume, and that is the whole point. At high throughput, the per-test premium on the "free" device dominates and the crossover comes early. At low throughput, wastage and fixed QC costs punish you instead. There is no single right answer, which is exactly why the sticker price cannot be the answer. The only honest way to compare two offers is to model the total cost of ownership across the realistic life of the contract, at your real volume, and to divide it by the number of results you will actually report.
The cheapest device on day one is very often the most expensive over five years. The only figure that settles it is cost per reportable result.
Lock-in: the cost you pay for having chosen badly
There is one more line that never appears in any model, and it is the switching cost. A placed analyser ties you to that vendor's consumables, its controls, its service and its data feed. When the contract renews, your negotiating position is weak, because leaving means new validation, new competency sign-off for every operator, a new EQA arrangement and a rebuilt connection to the record. Vendors know this. The generous day-one offer and the firm renewal are not a coincidence; the first buys the second.
This is why the shape of the contract matters as much as its price. A long exclusive term with punitive exit terms is worth less than a shorter one at a slightly higher price, because optionality has value and lock-in has a cost. Procurement that reads only the per-test figure and not the clause about minimum volumes, term length and exit is reading half the contract.
What to actually do
None of this argues against placement deals. Well-chosen, at the right volume, on the right contract, a low-capital placement can be genuinely the best value. It argues against choosing on the wrong number. Here is the practical discipline.
- Build a five-year total cost of ownership model before you decide. Capital or placement fee, consumables at realistic volume, controls, calibrators, EQA fees, service, connectivity, training and staff time. Put every line in, even the ones you have to estimate.
- Reduce every offer to cost per reportable result. This single figure is the honest comparator between a free device and a bought one, and it exposes the per-test premium that the capital price hides.
- Model your real volume, and stress it. Run the numbers at expected, low and high throughput. Wastage punishes the quiet clinic; per-test price punishes the busy one. Know which risk you are carrying.
- Ask the vendor the hidden-cost questions directly. What is the per-test price and does it change with volume? What is the on-board stability once opened? What controls and calibrators are mandated and at what cost? What EQA is required? What exactly does the service plan cover, and what does downtime cost me? What are the connectivity fees? What is the minimum-volume commitment, the term, and the exit?
- Read the contract shape, not just the price. Term length, minimum volumes, price-escalation clauses and exit terms are where lock-in lives. Prefer optionality; price it in.
- Count the staff time honestly. QC, maintenance, competency and EQA are recurring labour. A device that needs more of your people is more expensive, whatever the invoice says.
If you want to run this properly, our procurement support builds the model with you and sits on the other side of the table when the questions get awkward. We publish a procurement matrix template in our resource library so you can score competing offers on total cost rather than sticker price, and our training, including the free POCT Fundamentals course, covers the QC and competency costs that the model has to include. Understanding what each analyte actually demands in controls and EQA is part of the same discipline.
The number that matters
The next time a supplier slides a zero across the table, thank them, and then ignore it. Ask instead what a trustworthy result will cost you every single time you produce one, for the next five years, at the volume you actually run. That is the number that decides whether the deal is good. The sticker price was never the price. It was the bait.