By Varun Mahajan, FounderPublished Last updated Part of the pin code availability guide
A blank tells you less than you think
A product is missing from your report. Before you ask why it disappeared, ask what the report actually checked. Was the retailer read? Was the correct delivery area selected? Did the product match succeed? A missing answer can come from any of these steps.
Not checked means there is no supported answer for that product, location and time. Out of stock means the source was checked and showed an availability state. Treating those as the same result turns an incomplete comparison into a confident but unsupported claim.
Separate the five states
In stock means the source indicated that the product could be bought at the time and location checked. It does not guarantee that a later checkout will succeed. Availability can change between a page view and payment.
Out of stock means the source showed the product as unavailable at that check. Not listed means a completed check did not find the product in the defined set of listings. A failed search, a blocked page or an unmatched name does not establish either state.
Delisted describes a listing previously seen that is no longer found across the checks being compared. Keep the earlier record and the comparison dates. The change alone does not explain whether a retailer removed it, renamed it or changed its local range.
Not checked is the remaining unknown state. Record the reason when you know it. The retailer might be outside coverage, the location might not have been read, or the product might need a better match.
Keep the location and time attached
A result for one delivery area does not establish stock across a city. A result from yesterday is not a reading of the shelf right now. The product, source, location and time belong together when you interpret the answer.
Check these details before combining records. If one source has a current reading and another has only an older reading, show the difference. Calling both entries current hides a limitation the reader needs in order to decide.
Leave unknown values out of the stock answer
Imagine a worksheet containing a checked product that is unavailable and another product whose page could not be read. The first supports an availability conclusion. The second supports a coverage note. Entering zero for both loses that distinction.
Keep a separate measure of how much was checked. An availability percentage needs a stated denominator, the dates it covers and a definition of a valid check. Do not silently drop unknown records or count them as unavailable. Either choice can make a result look more complete than it is.
Work through a missing result
Start with the original product identity, including pack size and variant. Look for a renamed listing or a source-specific product reference. A similar name is not enough to decide that a replacement listing is the same item.
Next, confirm the location and the check time. Read the supplied source context and any recorded failure reason. If the evidence ends there, keep the state unknown. A clear explanation of what was not checked is more useful than a guessed answer.
Then ask whether that place was inside the dataset’s boundary at all. For Rewalto’s data, the data coverage page lists the pin codes held for prices and listings, with the date the catalogue was checked. A pin code outside that list was never read, which is a different state from a reading that failed.
When you repeat the check, keep the older result rather than overwriting its meaning. A later successful reading answers a later question. It does not prove what the source displayed during the earlier gap.
Stock data does not measure lost sales
An unavailable listing does not tell you how many people wanted to buy it, whether they switched to a competitor or what revenue was lost. Those conclusions need other evidence. A page can show a problem without measuring its commercial cost.
The same restraint applies to responsibility. A listing change does not identify who caused it. Use the result to decide what to investigate, then ask the team or retailer with access to the missing information.
Write the limitation where someone will see it
Put the stock state, location, check time and coverage reason beside the result. Do not bury them in a footnote while the headline calls the product absent everywhere. When a comparison mixes sources or dates, explain that before the reader draws a ranking.
A useful conclusion can be modest: the available checks show a gap in a particular area, while another area remains unchecked. That gives the next person a specific follow-up instead of a broad claim they cannot verify.
You do not need perfect coverage before using any data. A bounded result is useful when its limits are clear and fit the decision. A local availability check can support a local follow-up. It cannot support an all-city claim. The useful boundary is the question the evidence can answer, so write that boundary next to the result rather than in a footnote nobody opens.
Keep the result and the check outcome separate
A worksheet should make it possible to answer two questions independently: what did the source say, and did the check produce a usable answer? Keep the recorded stock value separate from the outcome of the check. A source that could not be read should not inherit the stock value from the previous day.
Keep a reason for an unknown result when it is available, but do not turn an explanation into evidence. Product not matched explains why a result is absent. It does not establish that the retailer has removed the product. If the reason was not recorded, say that the cause of the gap is unknown.
Watch what happens when you combine datasets
A join can create a blank even when the original source contained a value. Check your matching keys before interpreting a missing result as a source problem. Product references may be unique only within a particular retailer, and a location reference may not mean the same thing across sources.
Keep records that failed the join in a review list. Do not discard them silently. If a chart only includes matched products, label that population so readers know what the picture represents. It is a chart of the matched set, not necessarily the whole category.
Check the charting tool too. Some tools connect a line across a missing period; others draw a zero. Both choices can imply something the data does not say. Show the gap or explain the display choice, and make the underlying dates available to the reviewer.
Give the next person a testable follow-up
An account manager needs a specific question to take to a retailer. Name the product, location and period involved. Explain whether the evidence is a completed check, a failed match or a gap in coverage. Ask for the missing detail rather than presenting a diagnosis as settled.
A developer investigating a missing API result needs different context: the request filters, the expected record, the relevant source date and any receipt or check outcome. Share those through an appropriate private channel. Avoid sending access credentials or an entire customer file when a smaller description will do.
Once someone resolves the issue, update the explanation without rewriting the original source record. A corrected product match or a later successful check should remain distinguishable from what was known when the first report was prepared.
Choose your next step
Continue with same product different price across apps or check dataset coverage and freshness. Visit Rewalto to review current data and access options.
Common questions
Is "no data" the same as out of stock?
No. No data means the check did not produce an answer. Out of stock is a reading where the app showed the product as unavailable. Keep them in separate columns.
Why does the pin-code checker say "not collected here yet"?
It means Rewalto has no reading for that pin code and app yet, so the result is not evidence either way.
What should I do with unknown values in a report?
Show them as unknown, state how many there were, and do not average them into stock or price figures.
Can not listed and not checked be merged for a simple summary?
Keep them separate. They answer different questions, and the distinction is usually what the reviewer needs to know. If a presentation needs fewer categories, retain an explicit unknown group and explain how the checked results were grouped.
Does a product appearing again prove that a previous gap was a delisting?
No. A later appearance establishes a later reading. Earlier checks and their context are needed to describe what happened in between.