A tail spend threshold is a value ceiling per purchase order. Anything under it goes through a fast, light route; anything above it goes through normal sourcing. In the programmes we run, the ceiling is set per client at AED 50,000 or AED 200,000 per order, and it covers any category. The right number for you depends on how much your team can handle, how much risk you accept, and how fast requesters need to buy.
Why set the tail spend threshold per purchase order?
Most definitions of tail spend look backwards. They take a year of spend, sort suppliers by value and call the long list at the bottom the tail. That is useful for diagnosis, and useless for deciding what happens to the request that arrives this morning.
A per-order ceiling works at the moment of buying. A requester, a buyer or a system can look at one request and know which route it takes. Nobody has to know what the business spent with that supplier last year, or whether the category is "strategic".
It also replaces a lot of arguments. Instead of debating which categories belong in the tail, you agree one value and a short list of exceptions. Everything else follows from the order value.
Why not define the tail by category?
Category lists sound tidy, but they leak. Office supplies look like tail spend until someone orders furniture for a new building. Maintenance spares look strategic until the request is a single gasket for an old pump.
That is why the ceilings in our programmes cover any category, indirect or direct. A small one-off order for a production part needs the same treatment as a small order for printer toner: competed quickly, checked against the right suppliers, approved and recorded.
Keep a short list of categories that never go through the tail route, whatever the value. Typical candidates are things with safety, data or reputational risk that need a specialist eye. The list should be short enough to fit on one page.
What happens if the threshold is too low?
A low ceiling feels safe. In practice it pushes ordinary purchases into a process built for large ones.
- Your category team fills up. Buyers spend their week on small requests that do not justify a full sourcing event, and the strategic work waits.
- Cycle times grow. A request that could be done in days sits in a queue for weeks. Before METIS, two to three weeks was the usual cycle for this kind of purchase.
- People buy around the process. When the official route is slow, requesters use cards, expense claims or a supplier they already know. The spend still happens; you just lose sight of it.
- Orders get split. A ceiling that is clearly too low invites requesters to break one need into several smaller orders to stay under it.
What happens if the threshold is too high?
A high ceiling gives the fast route more volume, but it changes what that route must be able to do.
- The stakes per order rise. A larger order deserves more competition, closer reading of terms and a sharper look at supplier risk. A route designed for small purchases may not give it that.
- Contracted suppliers lose volume. If a category already has a negotiated agreement, large orders that bypass it weaken your position at the next renewal.
- Approvals still matter. Raising the ceiling should never mean skipping your own approval chain. It means the sourcing work is done differently, not that fewer people sign off.
The answer is not to avoid a high ceiling. It is to make sure the route below it can carry the weight: competed bids, scoring on more than price, supplier checks and an audit trail on every order.
How do you choose the right ceiling?
Start from your own data and your own capacity, not from a benchmark.
- Look at where your category team spends its time. Pull recent purchase orders and see how many small orders a buyer touched. If most of their week goes on orders well below their real expertise, your ceiling is too low.
- Check what sits just above your candidate ceiling. If those orders are already covered by contracts or frame agreements, the ceiling can sit lower. If they are one-off needs with no contract behind them, a higher ceiling usually serves the business better.
- Decide how much competition the fast route gives. A route that only places orders with one known supplier deserves a low ceiling. A route that competes every request, scores bids and negotiates can safely carry more.
- Agree the exceptions. List the categories that stay with your own team whatever the value, and who owns each.
- Set a review date. Treat the first ceiling as a starting point. Look at cycle times, savings and how often requesters still buy around the process, then adjust.
The two ceilings we use in practice reflect this. One programme runs at AED 50,000 per order; others run at AED 200,000. The difference comes from what each client already has under contract and how much their own team wants to keep.
What should sit below the line?
Under the ceiling, purchases still need routes. The usual split, covered in more depth in our tail spend management guide, is:
- Recurring, standard items go to a catalogue with agreed prices.
- One-off, clearly defined needs go to an automated quote process, so each is competed without staff time.
- Exceptions (unclear specifications, unusual suppliers, urgent requests) go to a buying desk with experienced buyers.
This is the structure METIS is built on: a catalogue for recurring items, AI sourcing agents for one-off needs, and a Buying Desk for exceptions. Below the client's ceiling, a requester describes the need in plain English; agents find suppliers, issue the RFQ, score bids on price, delivery, terms and supplier risk, and negotiate inside limits the client sets. The result goes through the client's own approval chain and the purchase order is issued with an audit trail. The aim is request to approved purchase order in around 48 hours. You can see the flow on how it works.
Next steps
- Pull recent purchase orders and sort them by value. Mark where your team's effort goes.
- Pick a candidate ceiling and look closely at the orders just above and just below it.
- Write down the categories that stay with your own team, whatever the value.
- Agree how orders below the ceiling are competed and approved before you switch anything on.
- Set a date to review the ceiling against cycle time, savings and off-process buying.
If you want to compare your numbers with what a managed route would deliver, the ROI calculator is a good place to start.