Ask your projects data anything. Get answers in seconds. Plexa MCP Connector is live. See it in action
FREE TOOL
Tender comparison and bid levelling tool
Compare subcontractor tenders on the same scope. Mark what each one included, excluded or left unstated, price the gaps back in, and see whether the ranking changes.
What bid levelling actually means
Levelling a tender means adjusting each price so that every tenderer is being compared on identical scope. Until you do it, you are not comparing prices at all. You are comparing prices attached to different amounts of work, which is a different exercise and a much less useful one.
The reason it matters is that exclusions are a pricing strategy. A subcontractor who excludes cranage, penetrations, testing and out-of-hours work will submit a lower number than one who includes them, and will be entirely honest about it in the fine print nobody read. The gap does not disappear; it reappears later as a variation, at a rate you are no longer in a position to negotiate.
Levelling puts those items back into the price at what you judge they would cost, so the comparison reflects the job rather than the paperwork.
Why the cheapest tender is so often the most expensive
The pattern is consistent enough to be a rule of thumb: on a package with five or more tenderers, the lowest submitted price is frequently not the lowest levelled price. Sometimes the difference is small. Sometimes it reverses the ranking entirely.
This happens because tendering rewards the appearance of being cheap. A subcontractor competing on price has two ways to lower the number: sharpen the rates, or reduce what the number covers. The second is easier, faster and lower-risk for them, and it is invisible unless someone reads the qualifications page and prices what is written there.
Awarding on submitted price alone means systematically selecting for whoever excluded the most. That is not a judgement about anyone’s honesty. It is simply what happens when the selection criterion is a number that means different things to different bidders.
The three things that turn into variations
Explicit exclusions are the easy case. They are written down, you can see them, and you can price them. The tool adds them back at whatever allowance you enter.
Provisional sums are the second case, and they are more slippery. A provisional sum is not a price. It is a placeholder that will be adjusted to actual cost. Carrying it at face value in your comparison assumes the allowance was right, which is exactly the assumption a provisional sum exists because nobody could make.
Unstated items are the third and the worst. When a scope item appears in neither the inclusions nor the exclusions, the tenderer has not made a commitment either way, and the argument about whether it was in the price happens later, on site, with the work already started. Silence is not inclusion. The tool carries unstated items at a risk weighting you set, and the right response to a long list of them is to send the question back before award rather than to price around it.
Why this tool has no built-in rates
Every free levelling spreadsheet that ships with benchmark allowances is guessing. What it costs to have someone else provide the crane, do the penetrations or run the testing depends on the market, the region, the programme, the trade and who else has capacity that month. A number that was right in Brisbane last year is not right in Perth this quarter.
The value of writing an allowance down is not that a tool produced it. It is that the estimator’s judgement becomes visible, dated and reviewable, so that when the variation arrives, there is a record of what was expected and why.
The one figure worth arguing about internally is the risk weighting on unstated items. Set it to zero and you are assuming silence means the item is included. Set it to a hundred and you are assuming silence means it is excluded. Neither is true, which is why the default sits in the middle and why the real answer is to eliminate the unstated items before you award, not to weight them.
What to do with the result
A levelled comparison supports a recommendation. It does not make one. Price is one input alongside capacity, licences, insurances, safety record, programme and whether the subcontractor can actually staff the job in the window you need.
Where the levelling changes the ranking, that is the finding worth writing down. It is the clearest possible evidence that the recommendation was made on scope rather than on the front page of a quote, and it is the thing a director or a client will ask about later.
Export the comparison to Word or CSV, attach it to the recommendation, and keep it. Six months later, when someone asks why the second-lowest tenderer was awarded, the answer should not depend on anyone’s memory.
The AI Construction Operating System
One platform, one data model, every project. Built in Australia, for the world.
YOUR DATA, PROTECTED
Hosted on AWS Sydney with Australian data residency.
GET THE MOBILE APP
Platform
Solutions
© 2026 Plexa Pro Pty Ltd. ABN 63 668 433 136