Price monitoring for retailers who have to defend margin

If you run ecommerce, merchandising or pricing at a mid-market retailer, competitor prices are not the deliverable. The deliverable is a defensible position on every product that matters, plus rules that cannot quietly spend your margin while you sit in a meeting. Monitoring supplies the inputs. Position and guardrails turn them into decisions.

The job is margin, not matching

Nobody is promoted for matching the lowest price on the internet. You are measured on gross profit dollars, on category growth, and on not being embarrassed on the twenty products customers actually price-check. A price feed does not resolve those tensions; it tells you faster that someone is cheaper. Treat it instead as a control system: decide where each group of products sits against the market, write that down as a rule, let monitoring flag the drift.

Price monitoring for retailers starts with the SKUs that carry the profit

Monitoring the whole catalog on day one is how programs die. Thousands of rows, no idea which deserve a response, a weekly export nobody opens by March. Start where a price decision changes the P&L, then widen.

A four-question score for choosing what to track

Score each candidate SKU from 0 to 3 per question. Track everything scoring 8 or more, review the 6 to 7 band once, leave the rest alone.

1. 1Gross profit share. Is it inside the products producing most of your gross profit? Profit dollars, not revenue, not units.
2. 2Price checkability. Can a shopper compare it in ten seconds elsewhere? A private label item nobody else stocks scores 0, however well it sells.
3. 3Competitive overlap. How many competitors you genuinely lose sales to stock it? One occasional seller is not overlap.
4. 4Margin sensitivity. How much does a five percent move change unit economics? A thin-margin accessory and a fixture react differently to one rule.

The daily loop: twenty minutes, four questions

A working daily routine is short and always the same. At an hour it gets skipped by week three, and a skipped routine is worse than none, because you still believe prices are being watched.

1. 1

   Is the data healthy?

   Coverage and freshness before any price. If a source degraded overnight, every number under it is stale and any rule reading it acts on fiction.

2. 2

   What moved on tier 1?

   Competitor moves on hero SKUs, with stock status attached. A competitor going out of stock at a low price is not a cut. It is a chance to hold or raise.

3. 3

   What is in the review queue?

   Every proposed change that hit a guardrail or a weak match waits here. Approve, edit or reject. An empty queue every single day means the guardrails are too loose.

4. 4

   Anything below a MAP floor?

   Resell under a MAP policy and your repricing must respect it. A competitor advertising below MAP is worth reporting to the brand rather than matching. See what MAP pricing is.

The weekly pricing review, and the decisions it must produce

The daily loop is maintenance. The weekly review is where strategy changes. Keep it to 45 minutes with merchandising, ecommerce and finance, and end every item with a decision, not an observation.

• Index by category, versus last week. Drifting down two points means your rules are following someone else's promotion. Decide: accept it, or tighten the band.
• Position mix. What share of SKUs sit lowest, matched or above? Decide whether that mix is the one you intended.
• The five most expensive changes. Not the most changes. The largest margin impact, each explained by a person.
• Competitor promotion cadence. Who ran what, and did it repeat? A competitor discounting the same category every third Thursday is a calendar, not an emergency.
• Floor hits and data quality. Any SKU that hit a floor twice, plus sources degraded and matches corrected. A repeated floor hit says the cost or the assortment is the problem, not the price.

Use price position instead of chasing every competitor cut

Chasing means answering each competitor move individually. It feels responsive and it loses: your most desperate competitor ends up setting your margin, and one seller clearing stock drags a whole category down in a fortnight. Position means deciding in advance where you sit, then acting only when reality leaves that band.

In practice, pick a target band per category, such as index 98 to 102 on core branded goods and 103 to 108 where you own delivery or installation, then let the rules hold it. When one competitor sits far below the band, that is a line to investigate, not a reason to move the category.

Two habits make position work. Weight by stock, because an out-of-stock competitor price is not an offer. And review the band quarterly: a target renegotiated every seven days is not a target.

Match the rule to the goal, not to the competitor

Most bad repricing comes from running one rule for four jobs. Protecting margin, buying share, clearing stock and defending a hero SKU need different metrics and rules. Mixing them is how a clearance ladder ends up cutting your best-known product.

Repricing guardrails, with the arithmetic shown

Guardrails are why automated repricing is safe to run. Four do nearly all the work, and each is a hard limit no rule can cross, whatever a competitor does.

• Margin floor. The lowest gross margin percent a SKU may reach, set per product or category.
• Cost floor. Landed cost plus a fixed adder for handling, returns and payment fees.
• MAP floor. Where you resell under a MAP policy, the rule stops at the advertised minimum instead of facing that brand's channel team.
• Max daily change. The largest move allowed in a day, turning a bad data read into a small mistake rather than a category-wide one.

Worked example. A tool kit sells at $89.00 and costs $58.50 landed, so gross margin is $30.50, or 34.3 percent. Set a 22 percent margin floor and the lowest price the rule may reach is 58.50 divided by 0.78, which is $75.00. A competitor drops to $71.90. The rule stops at $75.00 and files an exception, because following that competitor means 18.6 percent margin.

Add a 5 percent max daily change and the rule reaches $84.55 today, $80.32 tomorrow, $76.30 the next day and $75.00 on the fourth. The slowness is the point: four chances for a person to notice that the competitor price was a variant mismatch, a clearance unit or an out-of-stock listing.

Then the volume question, which decides whether the cut was worth making. Going from $89.00 to $75.00 is a 15.7 percent cut, and unit margin falls from $30.50 to $16.50. Earning back the same gross profit takes 30.50 divided by 16.50, about 1.85 times the units, meaning 85 percent more sales. Run that before the cut. The ecommerce margin calculator does it on your figures.

The phantom low price, and how it drains margin quietly

The failure mode that costs mid-market retailers most, described mechanically. Your rule matches to the lowest competitor. That listing is not comparable: a single unit where you sell a two-pack, a reseller with no stock, a refurbished item, last year's model under the same title. The rule cannot tell. It cuts. Nothing looks wrong afterwards, because the rule reports that it matched the lowest price, which it did.

Three defences stop it. Match on GTIN or UPC first, so identity is an exact join rather than a guess, and hold a confidence threshold below which a rule refuses to act and routes the change to review. Weight by stock, so an unavailable listing never sets your floor. Keep the daily cap tight, so a wrong read costs one step. See the product matching guide and repricing software.

What this will not do for you

• Not elasticity, and not motive. Those come from your own tests and your own read of a competitor.
• Not prices behind a login, and no checkouts are completed to compute a net price after cart coupons. Promotion text on the page is captured so a person can judge it.
• Not assortment, delivery promise or reviews. If you lose on those, a lower price only loses money faster.
• Not guesswork. Low-confidence matches wait in a queue, because a wrong match is worse than a missing one.

Matching, repricing, analytics, MAP monitoring, every integration and unlimited users are on every plan, Starter included. You pay for scale, never for a feature. See pricing and retail price monitoring.