Competitive Tracking That Produces Decisions
Most competitive dashboards are watched for a month and abandoned. What to track, how many, and the small number of questions the data can actually answer.
Competitive rank tracking is easy to start and hard to sustain, because the output is a chart and the requirement is a decision.
Start from the decisions
Write down what you would do differently depending on the answer. If nothing, do not track it.
Decisions rank data can inform:
Whether to enter a category. Whether a competitor's launch is gaining traction. Whether a price move worked. Whether demand in a category is growing or shrinking. Whether a competitor has a supply problem worth exploiting. Which competitor to benchmark against.
Decisions it cannot inform: anything about their costs, margins, strategy, or intentions.
Choosing the watch list
Small. Ten to thirty products for active tracking. A hundred-product dashboard is a dataset nobody reads.
Actual competitors, chosen by what a shopper would substitute, not by category membership. Marketplace categories are a seller's choice and do not define the competitive set.
Include yourself, obviously, and include a basket of neutral products for the control.
Include one or two category leaders even if you do not compete with them, as a market barometer.
Review the list quarterly. Competitive sets change and a stale watch list tracks products that stopped mattering.
What to capture
Rank, at a consistent cadence and time.
Price.
Availability.
Review count and rating, for the velocity proxy.
Seller, where visible.
Title, to detect listing changes.
Six fields. Rank alone answers almost nothing.
The weekly review
Fifteen minutes, and it is what turns collection into intelligence.
Any product outside its normal range, sustained.
Any availability change.
Any price change.
Any new product appearing in the tracked categories.
Basket movement, to identify market-wide shifts.
Anything unexplained goes on a list for investigation rather than into a conclusion.
Questions it answers well
"Is this competitor gaining or losing?" Over a month, with the basket, yes.
"Did their promotion work?" Compare their settled level after to their baseline before.
"Are they having supply problems?" Availability plus the decay-recovery signature.
"Is this category growing?" Basket median over quarters, with the caveat that catalogue growth confounds it.
"How big is the gap between us?" As a ratio of estimated units, with a range.
Questions it answers badly
"Why did they do that?" Never observable.
"What will they do next?" No.
"Are they profitable?" No.
"How much are they spending on advertising?" No, and estimates from placement observation are weak.
"What is their total revenue?" An estimate with an error range wide enough to be unusable for most purposes.
Presenting it
One chart with the tracked set and the basket median, inverted log axis, annotated with known events.
A short table: product, current rank band, direction over the period, availability, price, and any flag.
A paragraph of what changed and what it might mean, with the uncertainty stated.
Not a dashboard with forty metrics. The output of competitive tracking should be a page someone reads, not a system someone logs into.
Why these programmes fail
No decision attached, so the output has no consumer.
Too many products, so the review never happens.
No context fields, so every movement is ambiguous and the analysis stalls.
No basket, so market movement is read as competitive movement and the conclusions are wrong often enough to lose credibility.
No owner, so the collection breaks and nobody notices for a quarter.
The fix for all five is the same: decide what decision this informs, track the smallest set that informs it, capture context, keep a control, and give it to a named person.
The weekly note
The output that makes competitive tracking worth doing, and it is a short written note rather than a dashboard.
Five lines, weekly, to whoever makes decisions.
What moved, with the product named and the movement described relative to its normal range.
What explains it, or that nothing does.
What changed in availability or price across the tracked set.
Anything new in the category.
What to watch next week.
Written by a person who looked at the charts, not generated. The value is the judgement about which of the week's movements matter, and that is exactly what a dashboard cannot supply.
Keep the notes. After a year they are a narrative of the category that no chart provides, and they make the annual review straightforward.