Education procurement guide
A rollout is decided by two things: who holds the device, and how hard it works. Those answers set the durability, battery, serviceability and support model you need. Unit price is the last question, not the first. Below: the framework we use, a four-year cost model you can fill in, and the specification list to put in a tender.
Three schools can buy the same device and end up with three different four-year costs, because the cost is created after delivery, not at it. Damage rates, spare parts held, service logistics and staff time all move with how the device is used - and how it is used is a decision you make long before you shortlist a supplier.
That is why we start with deployment, not with a price list. If a supplier quotes you a unit price before asking who holds the device, they are pricing a product rather than a programme.
THE FRAMEWORK
Two questions decide most of the specification, and they are worth settling before you talk to anyone about unit price.
Cross those two and you get four education profiles. Every specification argument we have with a buyer is really an argument about which profile they are in.
Read down the column for your profile. The middle column is where we see money spent on the wrong thing most often.
| Profile | What actually decides it | Where money is usually wasted | Service model that fits | Platform |
|---|---|---|---|---|
| Cart-share or lab | Hinge and chassis life; port and screen durability; a fixed build across the cart | Paying for the largest battery and the fastest processor for browser-based work | Batch repair with screens and keyboards held locally | NB15 |
| Classroom 1:1 (on site) | Keyboard and screen longevity; one standardised image; manageability | Buying a unique configuration per year group instead of one build | Warranty plus a small on-site swap pool | M156FH |
| Take-home 1:1 | Battery capacity; chassis rigidity; offline-capable asset management | Trimming the chassis and the battery to protect the unit price | Formal swap pool plus a written family damage policy | S15 |
| Loaner and swap pool | Time to reimage; parts commonality; one image across every site | Per-site configurations that make pooling impossible | Parts held at the school or the district, standard build fleet-wide | All education platforms |
This is the section most unit-price comparisons leave out. The model below expresses everything as index points, where the device unit price is 100. That is deliberate: it means we are not quoting you a price, and you are not comparing our assumptions with another supplier's discount.
| Cost bucket (index points per device, 4 years) | Cart-share / lab | Classroom 1:1 | Take-home 1:1 |
|---|---|---|---|
| Acquisition: device and charger | 100 | 100 | 100 |
| Deployment: MDM enrolment, asset tagging, first setup | 6 | 9 | 12 |
| Consumables and spares held: batteries, keyboards, screens, chargers | 4 | 7 | 15 |
| Service and logistics: both directions, in and out of warranty | 5 | 8 | 17 |
| Loss and damage beyond economical repair | 1 | 3 | 12 |
| Programme administration and support desk time | 2 | 3 | 6 |
| Four-year total | 118 | 130 | 162 |
Read the top row against the bottom row. The device is identical in all three columns. The spread - 118 against 162 in this illustration - comes from spares, service, attrition and staff time, which is exactly where custody and duty show up.
Two things follow. First, a supplier who cannot tell you which of these buckets their offer makes worse is not helping you. Second, when you replace these inputs with your own, the ranking can change - which is why the model is worth building before you shortlist rather than after you have a preferred supplier.
Deployment cost in year one is not only hardware. It is imaging, asset tagging, printed guidance and staff training. If the platform underneath changes, that work is repeated. Asking a supplier to hold the mainboard, the exterior and the BIOS baseline means the year-one investment still applies in year two.
Tender documents tend to specify the product and leave the programme implied. These are the categories worth stating explicitly, because each one is a place a bid can look compliant and still fail in service.
| Category | What to specify | Why it decides the outcome |
|---|---|---|
| Deployment profile | State which of the four profiles the tender covers, and whether sites may differ. | It is the input to every other requirement. Without it, bids are not comparable. |
| Durability and mechanical | Hinge cycle expectation, port retention, screen impact resistance, and how durability is evidenced. | These are the failures that remove a device from service, and they are rarely tested in a one-week evaluation. |
| Battery and power | Required runtime under a stated load, charge-cycle expectation, and whether the battery is field-replaceable. | A non-replaceable battery effectively sets the device life. Ask before award, not in year three. |
| Software and management | Enrolment method, supported management tooling, image standardisation, and what happens offline. | Take-home fleets fail here first. |
| Documentation and compliance | The exact document set required for your market, as a named list, plus the battery transport documents (UN38.3 test summary and MSDS). | A vague clause is unenforceable at delivery. Name the documents, because a lithium battery cannot be shipped without them. |
| Service and lifecycle | Warranty start point and duration, parts availability window, swap or repair route, and turnaround expectation. | Determines buckets 3 and 4 of the cost model above. |
| Programme terms | Standard build across sites, repeat-order availability, and how discontinuation is notified. | Protects the year-one imaging and training spend. |
This is the part most suppliers leave until the evaluation meeting.
Your obligations are set by your market's regulator, your tender document and your channel's contract - not by a supplier page. These are the sources worth reading directly.
Send the deployment profile, the number of devices, the destination market and the timeline. We will come back with the configuration that fits, and with the gaps we can see.
Start a project