David Margaryan
Head of ASO at CheckASO
Where to hear and read about David Margaryan4 episodes and tests
App MastersApp Store Optimization for Million Dollar Apps (Part 4/4)
App MastersApp Store Optimization for Apps Ready to Scale (Part 3/4)
App MastersKeyword Prioritization for App Store Optimization (Part 2/4)
App MastersHow to Jumpstart Your App with App Store Optimization (Part 1/4)David Margaryan is Head of ASO at CheckASO who, across a four-part series, walks through App Store Optimization for an app's whole life: launch, keyword prioritization, scaling and running a million-dollar app. The lesson most worth taking from him is that ASO is not just keywords but a testing system tied to paid acquisition, ratings and creatives, where each channel feeds the others.
Everybody should use the paid marketing from the very start
How to Jumpstart Your App with App Store Optimization (Part 1/4)
David Margaryan: the story and the lessons
1Who David Margaryan is and how he defines ASO
Margaryan runs ASO at CheckASO, an app store analytics platform. He appeared across all four parts of an App Masters series on App Store Optimization, from jumpstarting a new app to optimizing million-dollar apps.
He often illustrates methods with CheckASO's own tools, such as its export of indexed keywords and its competitor finder.
Margaryan starts from store search because that is where most installs begin. Citing Apple, he says 65% of apps are installed from store searches, rising to 70% during the COVID pandemic.
Keywords make an app visible first to the store's algorithms and second to users, so keyword work comes before anything else.
He treats modern ASO as page improvement and constant testing, not just keywords. Graphic assets, text metadata, traffic channels and custom store listings can all be improved, and testing new hypotheses is what keeps already successful apps growing.
He calls it "a huge and complex task of improving the page and consistent testing of a new hypothesis."
2How Margaryan builds a keyword set for a new app
Competitor analysis is the first stage of every strategy he runs. He checks how rivals promote and whether search, explore or paid traffic leads, because if a category lives on brand keywords or paid promotion, text optimization alone may not reach the top.
Even then, better search visibility through text is a must in every strategy.
He mines competitor data as the main source of keywords. Popular apps are often indexed for a huge number of popular and uncommon keywords, so even a new, poorly indexed app can harvest them.
He calls competitor analysis "the main and the most effective way" to collect a semantic core. Tools speed it up, but the analysis is still yours to do.
Margaryan says you can often skip brainstorming today. Brainstorming queries from what the app solves became common when analytics services were poorly developed; now those services hold the data.
He checks which keywords an app is already indexed for. Apps are often indexed for relevant queries their owners never noticed. He puts those in the metadata and tracks them daily.
Store search suggestions come last, used only after every other source is exhausted, to make the pool a little larger.
Fight the competitors you share keywords with before the giants. A big keyword intersection marks your nearest competitors, who share your organic traffic, and CheckASO lists up to 20 of them based on that overlap.
Overlap is neither good nor bad: a meditation app with a high overlap with Calm can still compete.
Relevance is his first filter on any keyword list. Irrelevant keywords don't raise conversion or bring the right users, especially in the first stages of promotion.
A new app should climb from low frequency keywords to high ones. Top keywords belong to apps that have been on the market for years and promote on every channel. A new app goes after low and medium frequency, low competition terms, while a strong app can go after high competition ones.
Teams that load a launch with popular keywords fail to rank and then claim ASO doesn't work. "If you're a newbie, try to start with low frequency keywords, that's the right way to get higher."
He keeps weak and seasonal keywords on the list and tracks them. Popularity changes within weeks, and terms like New Year or Black Friday swing the most. The minimum Apple Search Ads popularity score is 5, which can still mean a little real traffic.
"Least popular doesn't necessarily mean zero popular."
Counting indexed apps is enough to judge competitiveness. Fewer apps indexed for a keyword means easier ranking. Difficulty scores blend popularity, competitiveness and relevance, and each tool calculates them differently.
He estimates difficulty himself from popularity and other available data rather than trusting one platform's score.
3Why Margaryan treats competitors' brand keywords as a risky extra
Branded keywords bring impressions but very few installs. People searching a brand already know which app they want, so even ranking second wins little.
In the same discussion, a client ranked second on Google for many competitor brands and saw no big lift.
Store guidelines forbid them, and enforcement varies by country. Using them can lead to rejection or, in extreme cases, suspension. Review is stricter in the US, while Margaryan says he can use any branded keyword in Russia without trouble.
A client whose US title used Periscope passed review but was then contacted by Twitter and told to stop.
Brand terms waste scarce App Store characters. They take space from a less popular but relevant generic keyword, so he uses them only when free space remains in the metadata.
Some brand keywords have very few apps indexed, so a new app can reach the top and catch some traffic. His answer: assess the risk and test it for your own app.
4Why Margaryan runs paid acquisition and ASO as one system
He wants paid traffic running from launch day. Store algorithms count total installs from all channels, so external traffic speeds up ranking and indexing, most of all in very competitive or very narrow niches and on Google Play.
Launch traffic also tests onboarding, retention and monetization, and early reviews show what to fix before scaling.
The more traffic you buy, the higher you rank, if it is relevant. Margaryan says thousands of tests by other ASO managers and by himself confirm the correlation.
Irrelevant traffic with poor engagement hurts rankings, so the source quality matters.
Size the starting budget from local cost per install. There is no fixed number: it can be $5 or a million. Multiply the average cost per install in your country by the installs you want.
Most developers lack paid acquisition experience, so he advises hiring experts, taking courses or reading up. Free channels like Reddit and TikTok save money but cost time and creativity.
He checks rivals' paid versus organic mix by hand. No tool shows the split with confidence, CheckASO included, and those that try rarely show their sources.
He looks at competitors' Facebook, YouTube and Google ad campaigns directly and treats Sensor Tower as a reference to verify.
Ad results tell him which keywords deserve organic placement. A keyword that converts well from tap to download in Apple Search Ads can be moved from the keyword field into the subtitle.
When a client's most relevant terms got no Apple Search Ads impressions, the team tested five keyword categories as Facebook interest audiences and used the best converter to guide ASO.
Ad creatives and store screenshots should be designed as one funnel. All ad traffic lands on the store page, so both must convert paid and organic users. Showing many real game levels in ads gave high conversion and good retention.
He weighs misleading gameplay ads against retention, since retention feeds back into keyword rankings. Ad headlines can inform screenshots, but he warns paid and organic users differ, so it sometimes fails.
Custom store listings match each campaign to its audience. Google Play allows up to five pages with different metadata per campaign, which lets users see a page in their language.
He expected Apple's custom pages, with a unique link per page for ad campaigns, to lift conversion from ad traffic.
5How Margaryan scales an app into new countries
Expand only after one country is stable and the next has demand. Stable means retention and lifetime revenue, not just installs and ranking. A snowboarding tricks app has no relevant traffic in Saudi Arabia or Turkey.
"If there is no traffic than any effort will be meaningless for you."
He relies on several briefed native speakers, not one. Awkward localized titles, common in the Russian store, inspire no trust and lose conversion, and a single native may be weak in the language.
He briefs them on what ASO is, the app, its goals and any brand or legal limits. A good native also catches wrong screenshot text, unpleasant colors and offensive wording.
Research local behavior alongside the native speakers. Competitor screenshots and ads, reviews, books, YouTube videos and UA creatives show what each culture responds to.
Freelancers on Upwork or Amazon Mechanical Turk can also be asked which native-language words they would search for an app.
Translate the app interface last. He localizes metadata first, tests the market, and translates the interface only once progress shows.
Some apps need it to enter at all: a children's fairytale app with localized metadata but no translated interface brings no relevant traffic.
6Margaryan's order for lifting conversion: ratings, then graphics, then extras
Ratings come before graphics once the text is tuned. About two thirds of users check the rating. A three-star app can lose up to half of installs, a one or two-star app up to 90%.
Graphic assets such as screenshots and icon come second, and tests of custom pages and promoted in-app purchases third.
Ask for a rating after a win, never mid-action. Good moments follow a finished hard level, a purchase or a completed order. Unhappy users share more readily, so timing matters.
Another trigger is opening the app 10 times in three days, which shrinks the audience but leaves users likely to rate positively.
Report fake competitor reviews and reply professionally. Competitors leave or buy negative reviews, which he sees often in Brazil, Russia and India. Reply, then report to store moderators explaining why the review is fake.
A genuinely low rating means the product needs work. It cannot be fixed in a day. Answer reviews, solve users' problems, and ask them to update their rating once resolved, since some will.
A chat channel with users softens negative reviews and shows what they don't understand and what features they want.
He says buying reviews does not work. Store algorithms recognize purchased reviews and remove them after a day or two, and in some cases the app is rejected.
Featured reviews, shown before the list expands, depend on length, number of ratings and app version, with reviews of the latest version rising to the top.
7What Margaryan tells million-dollar apps about ASO
Big apps skip ASO because budgets and rankings hide the gains. Large apps often rank first for key terms even with unoptimized metadata, since organic and external installs push rankings up, and a channel capped by store users looks small next to a big budget.
In the same discussion, a model shows the upside: at 1 million daily impressions and 5% conversion, a 5% relative lift adds about 2,500 installs a day.
Convince leadership with an audit and hard numbers. He runs a full audit of the page and store metrics and collects success stories from similar competitors.
If you can decide yourself, just start optimizing and let visible progress end the questions.
Getting featured is the most efficient next lever after the basics. Apply through each store's form with a real update, localization of metadata and interface, and use of new store features such as dark mode or live ops events.
Stores want to promote popular apps because part of user spending goes to them, and a feature can open new markets.
Prepare the app before the feature traffic arrives. Check server load, guidelines and stability, and submit the update well ahead with a detailed description, since moderation is slow.
Most featured traffic is not your audience, so conversion drops and ratings get mixed. Google Play weighs conversion heavily, so raise it beforehand.
Promoted in-app purchases let a top app hold two search slots. An Apple app can have up to 20, indexed by the same keywords as the app, so a first-ranked app's purchase often ranks second and pushes rivals out of view.
Each needs optimized graphics and text, and a tool can show which competitor purchases are promoted and indexed for which keywords.
App Store Optimization for Million Dollar Apps (Part 4/4)
Even apps with strong paid user acquisition still gain from ASO: small conversion lifts at large scale add up to big install and cost savings. The episode covers how to pitch ASO to leadership, how to prepare for getting featured, and why promoted in-app purchases are an overlooked ranking and conversion lever.
Margaryan runs ASO at CheckASO, an app store analytics platform. He appeared across all four parts of an App Masters series on App Store Optimization, from jumpstarting a new app to optimizing million-dollar apps.
He often illustrates methods with CheckASO's own tools, such as its export of indexed keywords and its competitor finder.
He treats modern ASO as page improvement and constant testing, not just keywords. Graphic assets, text metadata, traffic channels and custom store listings can all be improved, and testing new hypotheses is what keeps already successful apps growing.
He calls it "a huge and complex task of improving the page and consistent testing of a new hypothesis."
Big apps skip ASO because budgets and rankings hide the gains. Large apps often rank first for key terms even with unoptimized metadata, since organic and external installs push rankings up, and a channel capped by store users looks small next to a big budget.
In the same discussion, a model shows the upside: at 1 million daily impressions and 5% conversion, a 5% relative lift adds about 2,500 installs a day.
Convince leadership with an audit and hard numbers. He runs a full audit of the page and store metrics and collects success stories from similar competitors.
If you can decide yourself, just start optimizing and let visible progress end the questions.
Getting featured is the most efficient next lever after the basics. Apply through each store's form with a real update, localization of metadata and interface, and use of new store features such as dark mode or live ops events.
Stores want to promote popular apps because part of user spending goes to them, and a feature can open new markets.
Prepare the app before the feature traffic arrives. Check server load, guidelines and stability, and submit the update well ahead with a detailed description, since moderation is slow.
Most featured traffic is not your audience, so conversion drops and ratings get mixed. Google Play weighs conversion heavily, so raise it beforehand.
Promoted in-app purchases let a top app hold two search slots. An Apple app can have up to 20, indexed by the same keywords as the app, so a first-ranked app's purchase often ranks second and pushes rivals out of view.
Each needs optimized graphics and text, and a tool can show which competitor purchases are promoted and indexed for which keywords.
App Store Optimization for Apps Ready to Scale (Part 3/4)
Explains how to run ASO once an app is ready to scale: expand into new localizations, work on ratings, reviews and graphic assets, and connect ASO with paid user acquisition. Gives concrete tactics such as well-timed review prompts, fighting fake reviews, and feeding Apple Search Ads and Facebook learnings back into metadata.
Margaryan runs ASO at CheckASO, an app store analytics platform. He appeared across all four parts of an App Masters series on App Store Optimization, from jumpstarting a new app to optimizing million-dollar apps.
He often illustrates methods with CheckASO's own tools, such as its export of indexed keywords and its competitor finder.
The more traffic you buy, the higher you rank, if it is relevant. Margaryan says thousands of tests by other ASO managers and by himself confirm the correlation.
Irrelevant traffic with poor engagement hurts rankings, so the source quality matters.
Ad results tell him which keywords deserve organic placement. A keyword that converts well from tap to download in Apple Search Ads can be moved from the keyword field into the subtitle.
When a client's most relevant terms got no Apple Search Ads impressions, the team tested five keyword categories as Facebook interest audiences and used the best converter to guide ASO.
Ad creatives and store screenshots should be designed as one funnel. All ad traffic lands on the store page, so both must convert paid and organic users. Showing many real game levels in ads gave high conversion and good retention.
He weighs misleading gameplay ads against retention, since retention feeds back into keyword rankings. Ad headlines can inform screenshots, but he warns paid and organic users differ, so it sometimes fails.
Custom store listings match each campaign to its audience. Google Play allows up to five pages with different metadata per campaign, which lets users see a page in their language.
He expected Apple's custom pages, with a unique link per page for ad campaigns, to lift conversion from ad traffic.
Expand only after one country is stable and the next has demand. Stable means retention and lifetime revenue, not just installs and ranking. A snowboarding tricks app has no relevant traffic in Saudi Arabia or Turkey.
"If there is no traffic than any effort will be meaningless for you."
He relies on several briefed native speakers, not one. Awkward localized titles, common in the Russian store, inspire no trust and lose conversion, and a single native may be weak in the language.
He briefs them on what ASO is, the app, its goals and any brand or legal limits. A good native also catches wrong screenshot text, unpleasant colors and offensive wording.
Research local behavior alongside the native speakers. Competitor screenshots and ads, reviews, books, YouTube videos and UA creatives show what each culture responds to.
Freelancers on Upwork or Amazon Mechanical Turk can also be asked which native-language words they would search for an app.
Translate the app interface last. He localizes metadata first, tests the market, and translates the interface only once progress shows.
Some apps need it to enter at all: a children's fairytale app with localized metadata but no translated interface brings no relevant traffic.
Ratings come before graphics once the text is tuned. About two thirds of users check the rating. A three-star app can lose up to half of installs, a one or two-star app up to 90%.
Graphic assets such as screenshots and icon come second, and tests of custom pages and promoted in-app purchases third.
Ask for a rating after a win, never mid-action. Good moments follow a finished hard level, a purchase or a completed order. Unhappy users share more readily, so timing matters.
Another trigger is opening the app 10 times in three days, which shrinks the audience but leaves users likely to rate positively.
Report fake competitor reviews and reply professionally. Competitors leave or buy negative reviews, which he sees often in Brazil, Russia and India. Reply, then report to store moderators explaining why the review is fake.
A genuinely low rating means the product needs work. It cannot be fixed in a day. Answer reviews, solve users' problems, and ask them to update their rating once resolved, since some will.
A chat channel with users softens negative reviews and shows what they don't understand and what features they want.
He says buying reviews does not work. Store algorithms recognize purchased reviews and remove them after a day or two, and in some cases the app is rejected.
Featured reviews, shown before the list expands, depend on length, number of ratings and app version, with reviews of the latest version rising to the top.
Promoted in-app purchases let a top app hold two search slots. An Apple app can have up to 20, indexed by the same keywords as the app, so a first-ranked app's purchase often ranks second and pushes rivals out of view.
Each needs optimized graphics and text, and a tool can show which competitor purchases are promoted and indexed for which keywords.
Keyword Prioritization for App Store Optimization (Part 2/4)
Explains how to prioritize researched keywords for a new app's App Store and Google Play metadata, and whether to target competitors' branded keywords. The key lesson is to start with relevant, low and medium popularity keywords, and to treat branded keywords as a risky, low-install play.
Margaryan runs ASO at CheckASO, an app store analytics platform. He appeared across all four parts of an App Masters series on App Store Optimization, from jumpstarting a new app to optimizing million-dollar apps.
He often illustrates methods with CheckASO's own tools, such as its export of indexed keywords and its competitor finder.
Relevance is his first filter on any keyword list. Irrelevant keywords don't raise conversion or bring the right users, especially in the first stages of promotion.
A new app should climb from low frequency keywords to high ones. Top keywords belong to apps that have been on the market for years and promote on every channel. A new app goes after low and medium frequency, low competition terms, while a strong app can go after high competition ones.
Teams that load a launch with popular keywords fail to rank and then claim ASO doesn't work. "If you're a newbie, try to start with low frequency keywords, that's the right way to get higher."
He keeps weak and seasonal keywords on the list and tracks them. Popularity changes within weeks, and terms like New Year or Black Friday swing the most. The minimum Apple Search Ads popularity score is 5, which can still mean a little real traffic.
"Least popular doesn't necessarily mean zero popular."
Counting indexed apps is enough to judge competitiveness. Fewer apps indexed for a keyword means easier ranking. Difficulty scores blend popularity, competitiveness and relevance, and each tool calculates them differently.
He estimates difficulty himself from popularity and other available data rather than trusting one platform's score.
Branded keywords bring impressions but very few installs. People searching a brand already know which app they want, so even ranking second wins little.
In the same discussion, a client ranked second on Google for many competitor brands and saw no big lift.
Store guidelines forbid them, and enforcement varies by country. Using them can lead to rejection or, in extreme cases, suspension. Review is stricter in the US, while Margaryan says he can use any branded keyword in Russia without trouble.
A client whose US title used Periscope passed review but was then contacted by Twitter and told to stop.
Brand terms waste scarce App Store characters. They take space from a less popular but relevant generic keyword, so he uses them only when free space remains in the metadata.
Some brand keywords have very few apps indexed, so a new app can reach the top and catch some traffic. His answer: assess the risk and test it for your own app.
How to Jumpstart Your App with App Store Optimization (Part 1/4)
Explains how a new app should start App Store Optimization: build a keyword pool, mine competitors, pick low competition keywords first, and add paid traffic early. The result worth hearing is that store search drives 65% of installs (70% during COVID), so keyword work comes first.
Margaryan runs ASO at CheckASO, an app store analytics platform. He appeared across all four parts of an App Masters series on App Store Optimization, from jumpstarting a new app to optimizing million-dollar apps.
He often illustrates methods with CheckASO's own tools, such as its export of indexed keywords and its competitor finder.
Margaryan starts from store search because that is where most installs begin. Citing Apple, he says 65% of apps are installed from store searches, rising to 70% during the COVID pandemic.
Keywords make an app visible first to the store's algorithms and second to users, so keyword work comes before anything else.
Competitor analysis is the first stage of every strategy he runs. He checks how rivals promote and whether search, explore or paid traffic leads, because if a category lives on brand keywords or paid promotion, text optimization alone may not reach the top.
Even then, better search visibility through text is a must in every strategy.
He mines competitor data as the main source of keywords. Popular apps are often indexed for a huge number of popular and uncommon keywords, so even a new, poorly indexed app can harvest them.
He calls competitor analysis "the main and the most effective way" to collect a semantic core. Tools speed it up, but the analysis is still yours to do.
Margaryan says you can often skip brainstorming today. Brainstorming queries from what the app solves became common when analytics services were poorly developed; now those services hold the data.
He checks which keywords an app is already indexed for. Apps are often indexed for relevant queries their owners never noticed. He puts those in the metadata and tracks them daily.
Store search suggestions come last, used only after every other source is exhausted, to make the pool a little larger.
Fight the competitors you share keywords with before the giants. A big keyword intersection marks your nearest competitors, who share your organic traffic, and CheckASO lists up to 20 of them based on that overlap.
Overlap is neither good nor bad: a meditation app with a high overlap with Calm can still compete.
A new app should climb from low frequency keywords to high ones. Top keywords belong to apps that have been on the market for years and promote on every channel. A new app goes after low and medium frequency, low competition terms, while a strong app can go after high competition ones.
Teams that load a launch with popular keywords fail to rank and then claim ASO doesn't work. "If you're a newbie, try to start with low frequency keywords, that's the right way to get higher."
He wants paid traffic running from launch day. Store algorithms count total installs from all channels, so external traffic speeds up ranking and indexing, most of all in very competitive or very narrow niches and on Google Play.
Launch traffic also tests onboarding, retention and monetization, and early reviews show what to fix before scaling.
Size the starting budget from local cost per install. There is no fixed number: it can be $5 or a million. Multiply the average cost per install in your country by the installs you want.
Most developers lack paid acquisition experience, so he advises hiring experts, taking courses or reading up. Free channels like Reddit and TikTok save money but cost time and creativity.
He checks rivals' paid versus organic mix by hand. No tool shows the split with confidence, CheckASO included, and those that try rarely show their sources.
He looks at competitors' Facebook, YouTube and Google ad campaigns directly and treats Sensor Tower as a reference to verify.