How to Read Keyword Ranking History: Gaps, Missing Data and Scores

A keyword history chart helps you answer a practical question: what changed between the observations you can actually compare? Before acting on a rise, a fall or a blank space, check what the chart knows about each date.

LiteASO's keyword history makes missing days visible and separates observations that do not belong in one continuous series. This guide shows how to read that history, distinguish unavailable data from a checked result, and use summary scores without giving them more meaning than they carry.

Start with one comparable timeline

Open a tracked app's keyword workspace or keyword details, within your existing account limits. Pick one keyword and confirm the app, store and country before comparing dates. Keep the exact query in view: a similar phrase is a different question.

Then check the observation date and the available status. A recent page visit does not make an older measurement current. If you switch the scope or the underlying observations are incompatible, treat the sections as separate comparisons. LiteASO breaks those chart segments rather than drawing a line that suggests a continuous trend.

A useful first question is: “Am I comparing two measurements of the same thing?” If the answer is unclear, resolve that before calculating movement.

Read a gap as missing evidence

A blank day does not tell you the app held its previous position, fell to zero or disappeared from search. It tells you the timeline does not provide a comparable value for that day. Rank and demand charts preserve these missing days rather than filling them with a nearby value.

Imagine a hypothetical sequence: a rank of 18 on Monday, no available observation on Tuesday, and a rank of 14 on Wednesday. You can describe the two recorded positions. You cannot say the app climbed steadily across Tuesday, because Tuesday was not measured in that sequence.

The same reading applies to demand history. A missing point is not a recorded demand of zero. If you export notes or prepare a report, retain the gap and its date instead of smoothing it into a more reassuring line.

Separate unavailable from outside the checked results

“Unavailable” and “not found within the checked results” answer different questions. An unavailable observation cannot establish whether the app appeared in the results that a successful check would have examined.

A failed or stale rank observation must therefore not be read as evidence that the app is outside the checked results. Even a successful bounded check describes its checked scope; it does not establish absence from every possible result.

An older positive value cannot fill a newer unavailable value either. If Monday had a usable observation and today's observation is unavailable, Monday remains historical context. It is not today's measurement. When writing an update, say which date has the last usable observation and which date remains unavailable.

For a step-by-step investigation of an individual missing observation, use the missing keyword rank checklist. Here, the aim is to keep the whole timeline readable without inventing continuity.

Interpret the 30-day best in context

LiteASO's 30-day best uses comparable observations. Read it as the best supported position in that comparison window, alongside the history that produced it.

It does not mean the app held that position throughout the month. Nor does it fill gaps or turn an earlier best into the current rank. A useful reporting habit is to record the current available observation, the 30-day best and any missing intervals separately.

Suppose a hypothetical month includes a best observed rank of 9, a later rank of 16 and several unavailable days. “Best observed rank: 9” preserves the summary. “Current available observation: 16” preserves the later measurement. The missing days still deserve their own note; neither number explains them.

Keep scores, rank and confidence separate

The top-10 score is expressed on a 100-point scale. It is not a percentage, a probability of reaching the top 10 or a share of downloads. A score of 70 must not become “a 70% chance” in a report.

Metric confidence reflects the supporting evidence behind an estimate. New evaluations account for limited evidence more cautiously; older evaluations are not automatically recalculated.

Use each field for its stated purpose: rank records an observed position, the score provides its own point-based signal, and confidence helps you judge the estimate's evidential limits. None of these fields alone proves why a position changed or predicts a download outcome.

Turn the chart into a useful weekly note

For your next review, write down the keyword and scope, the dates of the comparable observations, the recorded movement, and any gaps or unavailable checks. Add the 30-day best as a separate summary and label the top-10 score in points.

Finish with the next check you intend to make. For example: “Compare the next available observation in the same scope before revising this keyword's priority.” This keeps your action tied to evidence you can inspect.

Keyword history covers the observations available for your tracked app and keywords. It is not a complete historical reverse index of every keyword an app has ranked for. Start with the tracked keyword you need to understand, keep missing evidence visible, and let the next comparable observation inform your next decision.

For the product changes behind this guide, see the keyword history and ranking clarity release notes.