Price to Win: Using Historical Award Data to Benchmark Contract Values

Most bidders price from the inside out: cost, overhead, margin, then a guess about what the buyer will accept. Award data lets you price from the outside in as well. It shows what similar buyers have actually paid for similar scopes, which is a more useful reference than a competitor's rumour or a buyer's estimate.

Award data is not a price list. It is a set of historical, imperfect observations. The value comes from cleaning it, banding it and comparing it carefully. This article explains how to do that and how to combine the result with your own cost floor.

What Award Data Can and Cannot Tell You

The buyer's estimate is usually confidential. The award value is public, but it may include VAT, escalation, options or multi-year totals that are not directly comparable. The discipline is to compare like with like and to be explicit about the adjustments.

The Fields You Need

The Tenders-SA awarded tenders

section and the procurement awards leaderboard

provide a starting dataset. The value estimator

produces a min, median and max range from historical awards for a category and province; use it as a first reference, then build your own comparable set.

Cleaning the Data Before You Trust It

1. Normalise VAT. Convert every value to the same basis. A VAT-inclusive award is about 15% higher than the same award excluding VAT.
2. Normalise duration. Convert multi-year values to an annual or monthly equivalent where the scope allows.
3. Remove obvious outliers. A value that is ten times the median may be a data error, a different scope or a bundled contract.
4. Check the scope. A "security services" award may include guarding, technology, monitoring and cash-in-transit. Compare the description, not only the category.
5. Adjust for time. Escalate older awards to current rand using a defensible index or a simple annual assumption, and state the assumption.
6. Separate provinces. Delivery cost and competition differ enough to distort a national median.

"

Building a Value Band: A Worked Example

A supplier is bidding for a 12-month office cleaning contract at a provincial department in Gauteng. It finds six comparable awards from the past three years, all excluding VAT and normalised to 12 months.

The median award is R2,800,000, between D (R2.55m) and F (R2.9m). The middle 50% of the data sits roughly between R2.55m and R3.1m. The per-square-metre figures range from R24 to R36, which is a wide spread and signals that the scopes are not identical.

1. Central reference: R2.8 million for 12 months, excluding VAT.
2. Plausible band: R2.5 million to R3.1 million, before adjusting for this tender's specific scope.
3. Scope adjustment: If this tender has 10% more floor area, the band might move to R2.75m–R3.4m.
4. Cost floor: If the supplier's verified cost is R2.65 million, the bid has room; if it is R3.2 million, the supplier is above the market band and must either reduce cost or accept a low-margin bid.

The supplier should not bid the median automatically. The median tells you where the market has settled, not where the winning bid will land. A competitive bid sits inside the band, above the cost floor and consistent with the evaluation system.

Price per Unit and the Comparability Trap

Per-unit comparisons are more useful than total values when the scope varies. For cleaning, price per square metre or per site; for security, price per guard post or per hour; for transport, price per kilometre or per delivery; for supplies, price per item.

The trap is that the unit may not capture the service level. A cleaning award at R24 per square metre may cover three cleans a week; another at R36 may cover daily cleaning, consumables and deep cleans. The per-unit figure is comparable only when the service level is comparable.

• Identify the unit that drives the cost.
• Check the service level behind each unit.
• Adjust for frequency, consumables, equipment and reporting.
• State the assumption in your pricing file.

The Bid Price Decision

The existing guide to tender pricing strategies

covers cost-plus, market-based and value-based models. The guide to B-BBEE preference points

explains how to adjust the competitive ceiling for your level.

When to Walk Away

• The market band is below your cost floor. Winning at a loss is not a strategy unless there is a deliberate, funded reason.
• A competitor is abnormally low. An abnormally low tender may be investigated or may fail in delivery. Do not match a price you cannot sustain.
• The scope has hidden costs. Access, security, consumables, reporting or penalty clauses can erase the apparent margin.
• The payment terms are too long. A profitable contract that requires three months of working capital can still sink the business.
• You cannot evidence the price. A price that depends on an unverified subcontractor or an unrealistic productivity assumption is a delivery risk.

The existing guides to abnormally low tenders

and loss-leader pricing

explain the risks in more detail. A low price can win the bid and lose the contract.

A Repeatable Benchmarking Routine

1. Pull 8–12 comparable awards from the last three years, same category and province where possible.
2. Normalise VAT, duration and scope, and remove outliers with a written reason.
3. Calculate the median, the middle 50% range and the per-unit figures.
4. Adjust for this tender's specific scope and service level.
5. Compare the adjusted band with your verified cost floor.
6. Adjust the competitive ceiling for your B-BBEE preference position.
7. Set the target price and record the assumptions in the bid file.

The value estimator

can generate the first range, and company intelligence

can show which suppliers win at that value. The existing guide to tender award data and market position

covers the strategic view.

Common Benchmarking Mistakes

• Comparing VAT-inclusive and VAT-exclusive values. A 15% distortion is enough to misprice a bid.
• Comparing different durations. A three-year total is not a one-year price.
• Ignoring the scope. The category label is not the scope.
• Using one award as the market. A single data point is an anecdote.
• Ignoring the date. An award from four years ago needs escalation.
• Bidding the median reflexively. The median is a reference, not a strategy.

A Comparable-Set Template

Keep the comparable set in a reusable spreadsheet so the next bid starts with a template rather than a blank page.

The confidence column matters. A high-confidence comparable has a similar scope, province, duration and service level. A low-confidence comparable is included for context only and should not drive the median. Weight the high-confidence rows more heavily, or calculate the band on them alone.

From Band to Bid Price

The band is not the bid. The bid is a decision inside the band, constrained by cost and strategy. A simple sequence works:

1. Set the cost floor: the lowest price at which compliant delivery is viable.
2. Set the market band: the range supported by the comparable set.
3. Set the preference-adjusted ceiling: how much price premium your B-BBEE position can absorb.
4. Choose a target inside the overlap of all three.
5. Stress-test the target against a 10% cost overrun and a 60-day payment delay.
6. Record the assumptions and get commercial sign-off.

If the three ranges do not overlap, there is no viable bid at the current cost base. That is a finding, not a failure. The response is to reduce cost, change the scope, find a partner or walk away, not to submit a price that cannot be delivered.

The band is a snapshot, not a permanent truth. Competition, input costs, buyer budgets and scope all move. Rebuild the comparable set for each major bid and date-stamp the result. A price model built on eighteen-month-old awards is a historical essay, not a competitive decision.

Use the model to set a range, then use judgement to choose a point inside it. The data tells you where the market has settled; it does not tell you what the buyer will accept tomorrow or what your competitor will do. Price-to-win narrows the guess, and the bid team still decides.

Conclusion: Let the Data Narrow the Guess

Award data does not remove the need for judgement. It narrows the range of plausible prices and exposes assumptions that would otherwise stay hidden. The best bidders combine three numbers: their verified cost floor, the historical market band and their preference-adjusted competitive ceiling.

For your next bid, build a comparable set before you open the pricing schedule. Clean it, band it, and write down the assumptions. Then price inside the band, above the floor and consistent with the evaluation rules. That is how you price to win without pricing to lose.