The Evaluation That Happens in a Spreadsheet
We can run a request for quotation and compare what came back on price. What we cannot do is score bids against weighted criteria, keep them sealed until opening, or record why the second-cheapest won — which is the part a grant-funded organization is usually required to be able to show.
There is a moment in competitive buying where the software usually stops and a spreadsheet begins. It is the moment the decision stops being about price.
Up to that point our procurement module does the work. You raise a [request for quotation](/glossary/request-for-quotation), send it to several suppliers, take their responses in line by line, and compare them. If the cheapest compliant quote wins, the process is complete inside the system and the trail is clean.
The moment somebody says the cheapest is not the best offer, everything after that happens somewhere else.
What exists, precisely
A request for quotation, and the quotations that answer it. That is the whole vocabulary. There is no separate notion of a request for proposal, where suppliers describe an approach rather than price a list. There is no request for information, the stage where you find out who exists before you ask anyone for anything. And there is no tender or bid as a thing in its own right.
- You cannot record evaluation criteria — the things you said you would judge on before you saw the responses.
- You cannot weight them, so "technical merit is worth twice price" has nowhere to live.
- You cannot score a response against those criteria, individually or as a panel.
- You cannot seal bids so that nobody sees them before a stated opening moment.
- You cannot record who evaluated, or keep two evaluators' scores separate from each other.
- And there is nowhere to write down why the winner won, other than free text on the resulting order.
The system can prove you asked three suppliers. It cannot prove how you chose between them.
Why that distinction is the whole post for a grant-funded organization
Most competitive procurement obligations that come attached to funding are not really about getting the lowest price. They are about being able to demonstrate, afterwards and to somebody sceptical, that the process you followed was the process you said you would follow. The criteria were set before the responses were seen. The people who scored were the people named. Nothing was opened early. The award matches the scores.
Every one of those is a statement about sequence and custody, not about arithmetic. And sequence and custody are exactly what a system can prove and a spreadsheet cannot — because a spreadsheet has no idea when it was written, who wrote which cell, or whether the criteria tab was edited after the bids arrived.
This is why the gap matters more for organizations working across the region on donor-funded programmes than it does for a trading company buying stock. A trading company that picks the second-cheapest supplier because the cheapest is unreliable has made a good decision and owes nobody an explanation. A programme office that does the same thing may be asked to evidence it two years later, by someone who was not there.
What the system can evidence about a competitive process
| Can you show it afterwards? | In the system | Only in a spreadsheet |
|---|---|---|
| That several suppliers were asked, and which | Yes | No |
| What each of them quoted, line by line | Yes | No |
| That the supplier was prequalified before bidding | Yes | No |
| That a blacklisted supplier was excluded, with the reason | Yes | No |
| That approval was obtained before the order was raised | Yes | No |
| The criteria you said you would judge on | No | Yes |
| The weight given to each criterion | No | Yes |
| Each evaluator's scores, kept apart | No | Yes |
| That no bid was opened before the opening date | No | No |
| Why the winner won | Partly — configurable by you | Yes |
Built and maintained Configurable by you, not maintained by us Not built
The sealed-bid row is marked as neither, deliberately. A spreadsheet cannot evidence custody either — it can only assert it. That row is the one thing on this table that genuinely needs software, and it is the one nobody has.
Read the left column and the honest summary is that we are strong on participation and absent on adjudication. Everything about who was invited, who was eligible, and who authorised the spend is recorded properly. Everything about how the choice was made is somewhere else.
What we would say to somebody in this position today
Use the system for what it evidences well, and stop pretending the spreadsheet is temporary. Run the request for quotation in the product so the invitation list, the responses and the approval trail are captured. Then run the evaluation deliberately and separately, with the criteria written down and circulated before the responses are opened, and attach the completed evaluation to the purchase order as a document.
That last step is worth insisting on, because it is the difference between an evaluation that existed and one you can find. An attached document on the order is not as good as structured scores — it cannot be reported on, compared, or checked for consistency — but it is durable, it is where an auditor will look, and it survives the departure of the person who ran the process.
Four questions for any system claiming to support competitive buying
Where do I record the criteria before I see the responses?
What a straight answer sounds like
A screen, or an admission. Ours is an admission.
Why it matters
Criteria recorded afterwards prove nothing, and everyone reviewing your process knows it.
Can two evaluators score independently without seeing each other?
What a straight answer sounds like
A demonstration with two logins.
Why it matters
Panel scoring that is really one person typing is a formality with extra steps.
Can a bid be opened early, and would I know?
What a straight answer sounds like
Ideally no, and yes. Realistically: no control at all.
Why it matters
This is the one requirement that genuinely cannot be met outside software, which makes it the sharpest question on the list.
Where does the award justification live, and is it reportable?
What a straight answer sounds like
A field, or an attachment. Ours is an attachment at best.
Why it matters
An attachment is findable. Free text on an order is not, and neither is a spreadsheet on somebody's laptop.
Three, and the order matters more than usual here
This is a well-understood shape and we would not invent a new one. What we would resist is building the scoring engine first, because scoring without the two things below it evidences nothing.
Evaluation criteria, recorded and locked before responses open
Criteria with weights, attached to the request, timestamped and unchangeable once the first response is received. This is the foundation: it is what converts an opinion about the winner into a comparison against something you committed to in advance.
Sealed responses with an opening event
Responses withheld until a stated moment, with an audit entry when they are opened and by whom. This is the requirement that cannot be met any other way, which makes it the one with the clearest case.
Independent scoring and an award record
Each evaluator scores without seeing the others, scores are combined by the weights already recorded, and the award captures the resulting ranking with any override stated as an override. Only worth building on top of the first two.
If a vendor offers you the third without the first two, what you are being sold is a calculator. The value in tender software is custody and sequence, not arithmetic.
Talk to us about tender evaluationWhat AWRA OpsHub does today
- Requests for quotation with per-item lines, and the responses captured against them for comparison.
- Supplier prequalification, so eligibility is decided and recorded before anybody is invited to quote.
- Vendor blacklisting with a reason, a date and the person who did it, which is an exclusion you can evidence.
- Requisition approval enforced before an order can be raised, so authorisation genuinely precedes commitment.
- Documents attached to the order, which is where an evaluation can live even though it cannot be structured.
What it does not do
- No request for proposal and no request for information. Only a request for quotation exists.
- No evaluation criteria, so what you promised to judge on cannot be recorded before the responses arrive.
- No weighting and no scoring, individually or as a panel.
- No sealed bids. Nothing withholds a response until an opening moment, and nothing records who looked at what and when.
- No award justification as a structured field — why the winner won is free text at best.
- No evaluation audit trail, so the sequence a reviewer would want to test is not captured.
Not ours, by choice
- We name no donor, agency or procurement rulebook anywhere in this post. What your funding requires of your process is between you, your funder and your auditor, and it is not a software vendor's to summarise.
- Prequalification is not evaluation and we do not present it as such. It answers whether a supplier may bid, not which bid should win.
The verdict
If you buy on price, our procurement module runs the whole process and the trail is genuinely good — invitation, response, eligibility, approval, order. If you buy on anything other than price, the decision itself leaves the system, and what you can evidence afterwards is that you asked several people, not how you chose between them. For a trading company that is a rounding error. For an organization whose funding comes with a process obligation attached, it is the part that gets examined, and you should plan for the evaluation to live in a document you attach rather than assume the system is capturing it.
Show us what you have to be able to evidence
The actual requirement, in the words your funder uses. We will tell you which parts we capture today, which belong in an attached document, and where the honest answer is that we are not the right system for that part of your process.
Talk to us about procurement