Review miningA comparison: ASO and keyword tools
What is in this resource:
How to build your ASO strategy from end to end
Learn to pick target keywords, write metadata, follow rank and base every decision on data.
From data to action
Simplify the keyword data and stay in line with App Store rules.
Tools that lift rank
Carry the keywords drawn from competitor reviews into your metadata and take your app out of the cusp band into the top 10.
The question that replaces a feature grid
The only honest way to evaluate an ASO tool is to ask where each number it shows came from. Measured, derived, or estimated? That one question surfaces the real difference between tools.
ASO tools get compared by putting feature lists side by side: this one has keyword tracking, that one has rival analysis, the third has automated reports. The list says nothing, because very different work hides under the same heading.
The question worth asking is where a number came from. Three answers are possible and they are not alike.
It may come straight from Apple: the rank in a search result, category charts, app metadata, ratings, reviews. Those are public and verifiable - you can search and see the same thing.
It may be derived from Apple's signals. Popularity works that way: a relative value obtained from search suggestions. Not an absolute number, and it does not claim to be.
Or it may be estimated from a model: download count, revenue, a rival's keyword field. Apple does not publish those; the figures given are computed backwards from rank.
The single thing to check when evaluating a tool is whether the numbers in the third group are clearly marked as estimates. If they are not, the user takes an estimate for a measurement and decides accordingly.
What Apple gives openly and what it does not
That distinction is the foundation of the whole comparison, so let it be written plainly.
What Apple gives openly: search results and ranks, category rankings, app name, subtitle, description, price, category, rating score and count, review text, search suggestions.
What it does not give: download count, revenue, conversion rate, absolute search volume, an app's keyword field, and the components and weights of the ranking algorithm.
Those two lists are the same for every tool. None of them has a private data arrangement with Apple; they all read the same open sources.
So the difference between tools is not in access to data. It is in two other places: coverage and honesty.
Coverage is how many countries are measured, how often, and how long the history is kept. Honesty is whether what could not be measured is presented as if it had been.
| Figure | Its source |
|---|---|
| Rank | Apple - read straight from the search result, verifiable. |
| Rating score and count | Apple - public on the app page. |
| Review text | Apple - public. |
| Popularity | Derived from Apple's search suggestions; relative, not absolute. |
| Download count | Estimate - computed backwards from rank. |
| Revenue | Estimate - derived from the download estimate; error twice over. |
| A rival's keyword field | Estimate - inferred from rank data. |
Coverage: how many countries, how often
This is the most concrete difference between tools, and it is measurable.
Country count. Apple's store is open in 175 countries and regions. How many does a tool measure rank in? That question can be asked directly and the answer verified: pick a country and try to track a phrase there.
Measurement frequency. Daily or weekly? Without rank history no move's effect can be measured, and weekly measurement can miss a metadata move's arrival entirely.
The unit of measurement. Is measurement per phrase or per row? That technical detail goes straight into the price: in a system measuring per row, adding a rival costs money; in one measuring per phrase it does not.
How long history is kept. Three months is not enough to see seasonality. A year lets you compare against the same period last year.
How an unmeasured value is displayed. Is a phrase whose popularity could not be measured shown as zero or as blank? A tool showing zero drops that phrase to the bottom of the list and you take it for worthless.
Honesty: saying which figures are estimates
The second difference is less visible and more important: whether a tool states its own limits.
The most common overreach is the download estimate. When a tool says "this app gets 40,000 downloads a month", that number was computed backwards from rank and rating count. The error bar varies widely by category and is usually not printed.
The second is a rival's keyword field. When a tool says "your rival's keyword field is this", that field is not shown by Apple; an inference was made from rank data. The inference is sometimes right and often incomplete.
The third is absolute search volume. "This phrase is searched 12,000 times a month" cannot be said, because Apple does not publish volume. Relative popularity can be given; an absolute number cannot.
The fourth is showing an unmeasured value as zero. That is the most insidious: zero is a reading and people decide on readings. If it could not be measured it should be left blank.
You can test those four points during a trial. Look at a screen asking for something unmeasurable and see what stands there: a number, or the words "could not be measured"?
- Every tool showing downloads and revenue is estimating; does it say so?
- Every tool showing a rival's keyword field is inferring.
- Absolute search volume cannot be given; relative popularity can.
- An unmeasured value should carry a blank, not a zero.
The misleading part of a feature list
Very different work happens under the same feature name. Four examples show where the difference lies.
"Keyword tracking". In one tool that means recording the daily rank of phrases you choose. In another it means listing the phrases but measuring rank weekly. In a third it means never measuring rank at all and only showing popularity.
"Rival analysis". In one it means reading your rivals' ranks for your tracked phrases out of the same search. In another it means showing their metadata. In a third it means showing their estimated downloads - which is an estimate.
"Keyword suggestions". In one they come from search suggestions. In another from rival metadata. In a third from rival reviews - and that last one is the place the others do not look.
"Metadata audit". In one it means counting characters and listing the waste. In another it means giving an overall score - and a score does not tell you which job to do.
So feature lists are not compared; the question under each feature is asked: where does this number come from, and how often is it measured?
What the pricing model reveals
A tool's pricing model gives away how it set up its measurement, and that determines what you will pay over time.
Per-row pricing. If every app-phrase-country combination you track is counted separately, adding a rival costs money. That means measurement runs per row: a separate search for each row.
Per-phrase pricing. If every app tracking the same phrase is read out of one search, adding a rival costs nothing. What raises cost is the number of phrases and the number of countries.
The second model is both cheaper and more accurate: your rank and your rivals' are read at exactly the same moment from the same result list, which makes them comparable.
Per-country pricing exists too and it is reasonable: each country means searching all of your phrases again. Adding a country really is expensive work.
The model to avoid is the package where it is unclear what you are paying for. A promise of "unlimited keywords" means either the measurement frequency is low or there is a limit somewhere - and both surface later.
Ten questions for a trial
Questions to ask during a trial. The answers should be findable in the interface or the documentation; if they are not, that is an answer too.
- How many countries does it measure rank in?Apple is open in 175. Easy to test: try to select a country that is not on the list.
- How often is rank measured?Daily or weekly? Weekly can miss a metadata move's arrival.
- How long is history kept?Three months does not show seasonality; a year allows a same-period comparison.
- How is an unmeasured value shown?Zero or blank? That one screen shows how the tool thinks about data.
- Are downloads and revenue marked as estimates?Apple publishes neither. Unmarked, a user takes an estimate for a measurement.
- Is a rival's keyword field shown?If it is, an inference is being made. Is it stated as one?
- Does adding a rival raise the price?If yes, measurement is per row; if no, per phrase. The second is cheaper and more accurate.
- Is there review data?Few tools extract phrases from rival reviews, and users' own words are in no other pool.
- Can you export the data?The measurement history is yours and should be portable; one that cannot move is reset when you switch.
- Is there an API or a command line?Can you connect your own measurement to your own processes? It shows whether the tool is a closed box.
What switching costs
Moving from one tool to another costs more than it looks, and most of the cost is on the data side.
The first loss is rank history. The new tool starts measuring from zero. If the old one's history can be exported you can carry it; if not, months of accumulated comparison points are deleted.
The second is a difference in method. Two tools may normalise the same phrase slightly differently - especially if their letter-folding rules differ. Then old and new measurements cannot be compared directly.
The third is measurement time. If one measures in the morning and the other in the evening, an artificial jump appears on the switching day. Writing that jump down to a move is the most common switching mistake.
The fourth is the pool. Rival lists, tracking lists and tags usually cannot be moved and have to be rebuilt by hand.
So running both for a month during the switch makes sense. Comparing what the two show for the same phrases both measures the difference and validates the new tool.
One last warning
Tool choice is the most discussed and least decisive decision in ASO.
What is decisive is measurement discipline: measuring before a move, changing one field per release, subtracting the control group, and keeping a record. Without those four no tool works.
With them in place, the difference between tools comes down to time saved - a real gain, but not the decision itself.
So the most important question while trialling a tool is this: does it make measurement discipline easier or harder? One that makes record-keeping easy, shows the control group and leaves unmeasured values blank is supporting the discipline.
Frequently asked
- Which tool has the best data?
- None has a private arrangement with Apple; they all read the same open sources. The difference is coverage - how many countries and how often - and honesty, meaning whether estimates are labelled.
- How reliable are download estimates?
- They vary widely by category and the error bar is usually not printed. They are computed backwards from rank and rating count; a decision should not rest on them.
- Can I really see a rival's keyword field?
- No. Apple keeps it private to the developer. Tools claiming otherwise infer from rank data; the inference is sometimes right and often incomplete.
- Do free tools do the job?
- For one country and a handful of phrases, yes - it can even be done by hand. The limit is daily measurement across 175 storefronts and keeping the history; both need automation.
- What happens to my history if I switch?
- If you can export it, it moves; if not, it is deleted. Running both for a month during the switch measures the difference and validates the new tool.
Explore our resources
We explain rankcusp's concrete return step by step
Discover the direct return of search visibility. This resource explains the effect on downloads of carrying cusp band keywords into the top ten, why phrases drawn from competitor reviews bring cheap traffic and how country-level prioritisation protects your budget. With rankcusp, ASO stops being a maintenance task and turns into a measurable growth channel.
See resource
rankcusp explains: the latest changes in the App Store search algorithm
Be ready for the changes in the App Store's search algorithm. Download the short guide explaining what has changed recently in store search: which metadata field gained weight, who is affected, what is at risk. Whether you are publishing your first app or growing your catalogue, discover how we make your job easier with daily rank tracking, the cusp band report and ready metadata suggestions.
See resource
Keyword research - the road map to the top ten
Want to get ahead in a category where the competition has tightened? Download our free report "Keyword research: the road map to the top ten" and see what measuring the words in the cusp band, reading the search language in competitor reviews and building the 100-character keyword field properly gains you over the long run. Learn the applicable steps that will strengthen your metadata and win a lasting place in search. Do not miss it - download the full guide now and rebuild your keyword strategy today!
See resource