One furniture retailer drops the price of a handful of dining sets by 20% on two weekends. Another keeps half its comparable range 8% below its usual price for three weeks in a row. The first retailer makes the larger cuts. The second gives shoppers more opportunities to find a discounted set. If your category review records only the biggest saving, you lose that distinction.

A competitor promotion analysis tracks how often reductions appear, how far prices fall and how widely the reductions reach. Start with one category, use the same products and observation dates throughout, and report each measure alongside its sample size. That gives a furniture team a practical way to compare recurring patterns before deciding whether its own category warrants a closer look.

Build a sample you can follow for several weeks

Choose a category narrow enough that the products are plausible alternatives. For dining furniture, you might select six-seat indoor sets with a table and six chairs, then group them by material and usual price band. Put tables sold on their own and sets that swap in a bench outside that comparison. Our comparable furniture pricing guide explains how to record the differences that matter to a buyer.

Within each retailer, keep one row per selected product and configuration. A dining set listed separately in four colours can dominate the results if all four appear as separate products. Choose a representative finish before you start or record variants separately when differences between finishes are the question. Keep the same treatment for every retailer.

Select your products before looking for discounts. A sample built from sale pages will naturally have more discounted products than a sample built from the category as a whole. Include a spread of relevant price bands and keep a record of the original list. If one set disappears, keep its row and mark later observations missing. Replacing it with a freshly discounted product midway through the period makes a retailer look more promotional even if nothing changed in the original sample.

Pick the same observation days and approximate times for each retailer. To compare weekly exposure, for example, take a snapshot at the same time each Thursday. Add a second fixed checkpoint if weekend offers matter and you have the resources to check both retailers consistently. Treat any event weeks separately if the aim is to understand ordinary trading. A dedicated Black Friday checklist covers the different decisions that arise around a specific sale event.

Decide what counts as a comparable reduction

Record the actual displayed product price, the configuration it buys and the date you observed it. Keep the site's advertised saving in a separate field. For the analysis, set a reference using earlier credible observations of the same configuration. A stable observed price across several checks is a stronger reference than a crossed-out figure you have never seen charged. Record the reference window and refresh it when later evidence supports a lasting price change. The competitor price history guide covers how to distinguish a temporary move from a new baseline.

For the main comparison, count a product as reduced when you can verify a lower publicly displayed price for the selected configuration against the observed reference. Use separate flags for public codes, member-only offers, bundles or discounts that depend on basket size. If those offers are important, review them as a second set under consistent eligibility assumptions. For example, a free-chair offer can matter to a shopper buying a complete room, but its value belongs in a bundle comparison with the same contents.

Use explicit observation states:

StateExampleTreatment
ReducedSame oak table and six chairs observed at £720 against a credible £900 referenceCount the reduction and record its depth
At or above referenceSame set observed at £900 or moreInclude as evaluable with no reduction
UncertainSet observed at £720 with a “was £900” label and insufficient earlier observationsKeep the observation, leave reduction status unresolved
MissingThe page was inaccessible during the scheduled checkKeep the gap and exclude it from evaluable observations

A Fido product price history can contribute observations for monitored products. Checking offer conditions, choosing reference prices and making the category comparison remain part of the team's review.

Calculate three measures with visible denominators

Retail pricing research treats promotion frequency and depth as separate measures. Adding sample coverage helps distinguish a discount on a few products from one spread more broadly through an observed range. (Circana, Hitsch, Hortaçsu and Lin)

Frequency — how often did any reduction appear? For each retailer, count scheduled snapshots with at least one verified reduction and divide by the number of snapshots that meet your evidence rule. Agree the rule beforehand. With ten products per retailer, you might require at least eight evaluable observations for a snapshot to qualify and use only dates that qualify for both retailers. State the rule alongside the result. If a reduction appears on three consecutive Thursday checks, report three observed dates with a reduction. Count a separate promotional episode only when evidence distinguishes it from the last one. A gap with no observations leaves the episode count uncertain.

Depth — how far did the price fall when reduced? Subtract the observed reduced price from its reference and divide by the reference. A move from £900 to £720 is £180 off, or 20%. For the category summary, report the median across reduced product observations and give the observation count. Retain the pound amount for products that matter commercially. Using repeated observations means a sustained reduction contributes once per qualifying snapshot. For a review of distinct campaigns, build an episode-level summary and label it accordingly.

Coverage — how much of the sample was reduced? For each snapshot, divide the number of reduced products by the number with evaluable observations. Five reductions out of ten evaluable sets gives 50% observed coverage. If three sets were missing or uncertain and the other seven include five reductions, you observed five out of seven, with three unresolved. Under the example eight-product rule, exclude that date from the retailer comparison and retain the underlying observations for follow-up. Where sample availability differs, give the raw counts beside any percentages.

Keep the measures side by side. Frequency answers whether a shopper checking at the selected times regularly encountered some form of reduction. Depth gives the typical size of the observed reductions. Coverage shows the share of the evaluable sample discounted at each checkpoint. A retailer with a larger sample has more chances to show at least one reduced item, so use comparable sample sizes and review the coverage figures before drawing conclusions from frequency alone.

Worked example: ten dining sets, four Thursday checks

The figures below are hypothetical. Both retailers have ten evaluable sets at each check, every discounted observation at Retailer A is 20% below its reference, and every discounted observation at Retailer B is 8% below its reference.

MeasureRetailer ARetailer B
Reduced sets at each check2, 0, 2, 00, 5, 5, 5
Checks with a reduction2 of 43 of 4
Median depth on reduced observations20% across 4 observations8% across 15 observations
Peak observed coverage2 of 105 of 10
Reduced product observations across the period4 of 4015 of 40

Retailer A's larger reductions occurred on fewer checks and reached fewer sets at a time. Retailer B's shallower reductions were visible across more of the sample on three consecutive checks. The recorded dates leave open whether B's offer ran continuously between checks. They also leave open whether A's first promotion ended immediately after the first check or closer to the second.

There is a second question before either pattern becomes actionable. If Retailer B's discounted products still cost more than your closest equivalents, the coverage result can help you understand what shoppers see without suggesting a price change. Check actual price levels separately. A competitive price index can compare a deliberately chosen basket of comparable offers on the same date. The promotion analysis describes the pattern across dates.

Bring one category decision to the review

Start with the part of the range where the observation and your own evidence meet. If repeated reductions sit mainly in premium oak sets while your overlap is in entry-level veneer, keep the two positions separate in the review. If the same competitor has several comparable oak sets reduced across multiple checks, examine whether your products have a relevant difference in specification, availability or the customer proposition. Use your own sales, stock and margin to judge what you can test and what you can afford.

Bring a short record to the meeting:

  • Sample and dates — category, selected configurations, price bands and observation schedule.
  • Evidence quality — counts of evaluable, uncertain and missing observations, plus excluded dates.
  • Three measures — frequency, median depth and coverage by check, with their denominators.
  • One action — for example, verify the oak-set price position before the next category review, or investigate whether product pages explain the extra chairs and finish options clearly enough.

Public prices describe the offers you observed. The retailer's units sold, margins, promotional return and reasons for discounting require other evidence. Set your own prices independently and keep future pricing decisions within your own team. CMA guidance for online sellers

For your next review, take one stable category sample and carry forward the same observation rules. The useful comparison is whether reductions are appearing more often, reaching more relevant products or cutting deeper than in the previous comparable period. That record can show which part of the category merits a closer look.