Marketplace App Marketing: How to Grow Both Sides of a Two-Sided App

Marketplace app marketing is the practice of acquiring, activating and retaining both sides of a two-sided app: the buyers, renters or shoppers who create demand, and the sellers, landlords, agents, retailers or service providers who create supply. A two-sided marketplace is a platform whose value to each user depends on how many relevant users exist on the other side. That single fact changes how a marketplace app should be measured, targeted and budgeted, because an install on one side is only worth something if the other side is there to meet it.

Marketplace liquidity is the probability that a user who arrives with intent finds a good match quickly: a flat to view, a deal to save, a policy to compare, an order that arrives on time. This guide explains how Admiral Media plans paid user acquisition for marketplace and listings apps, with real results from Admiral Media’s work with ImmoScout24, kaufDA, Clark and a UK fast commerce app, plus the platform features that matter most when one app has to speak to several audiences at once.

Admiral Media has managed more than €500M in ad spend for 150+ mobile brands. The patterns below come from that work, and every number in this article is linked to its source.

What makes marketplace app marketing different from marketing a single-sided app?

Why can’t a marketplace app be marketed like a game or a subscription app? Because a marketplace sells access to other people’s inventory, so the product the user experiences is only as good as the supply in their segment, their city and their price range. Paid media that ignores the other side of the market buys installs the product cannot satisfy.

A fitness app ships the same workouts to every user. A property app ships whatever listings exist in the user’s search area today. That dependency creates four problems single-sided apps rarely face.

• One app, several audiences with different intent. A real estate app serves people who want to rent and people who want to buy, and often agents and private owners as well. Each group searches for different words, responds to different promises and converts on different events.
• The install is a weak signal. In a marketplace the value appears at the first meaningful interaction with supply: an enquiry, a saved search, a lead, a first order. Optimizing to installs teaches the ad algorithms to find people who download and leave.
• Supply quality shapes acquisition efficiency. When a segment has thin inventory, conversion rates fall and acquisition costs rise, no matter how good the creative is. Paid media can expose a supply gap faster than any dashboard.
• Trust carries more weight. Users hand over data, money or a phone number to strangers through the platform, so safety and real inventory matter as much as price.

Venture investor Bill Gurley described why some marketplaces become large and others stall in his essay All Markets Are Not Created Equal, which lists factors such as supplier fragmentation, purchase frequency, network effects and the friction of signing up suppliers. Those same factors decide how a marketplace should spend on user acquisition. A high-frequency marketplace can pay for a first transaction because repeat use follows; a low-frequency one, such as property, must win the user at the moment of intent.

Which side of a marketplace should you acquire first?

Which side should get the budget first? In most marketplace apps Admiral Media works with, paid media should go to the side that is harder to keep engaged, while the supply side is checked for depth before any demand campaign scales. Spend follows constraints: whichever side is scarce in a given segment is the side that limits growth there.

An established listings app with millions of listings is rarely supply constrained at the national level. Its growth problem is converting searchers into active users, so paid media should go almost entirely to demand, segmented by intent. A young local marketplace is the opposite. It may have plenty of interested buyers but too few providers in most neighbourhoods. Buying more demand there makes the first experience worse.

The useful question is not “supply or demand?” but “where is the constraint, segment by segment?” The Admiral Media team usually answers it with three checks before scaling any campaign:

• Search to result rate. What share of searches in a segment return enough relevant results to act on? Low rates point to a supply gap.
• Time to first match. How long does a new user take to reach the first enquiry, lead or order? Long delays usually mean either thin supply or a confusing onboarding.
• Repeat behaviour by segment. Which cohorts come back? Segments where users return are where demand spend compounds.

Many marketplaces acquire supply through sales teams or listings feeds rather than app installs. The paid team still needs that supply data, because it decides which geographies and keywords deserve budget.

How do you choose the right activation event for a marketplace app?

How do you pick the event to optimize toward? Choose the earliest event that reliably predicts a transaction or a returning user, and that happens often enough for ad platforms to learn from. In marketplaces that is usually the first meaningful interaction with supply, not the install and not the final purchase.

Ad algorithms learn from the events you send them. Too shallow, and they find cheap installers; too deep and rare, and campaigns stay stuck in learning. Find the middle event, then move deeper as volume grows. Admiral Media’s guide to app event optimization explains the mechanics of this event ladder in more detail.

The table below shows the activation events Admiral Media typically evaluates for different marketplace models. It is a decision aid, not a benchmark: the right event for a specific app depends on its funnel data.

Admiral Media’s work with Clark, a German insurance app, shows why the middle of the funnel deserves attention. The case study notes that sometimes it is better to create your own event rather than push general events where bids are higher, and that optimizing for events that “stand on the sideline” of the main funnel helped performance, because users explore the app before entering the purchase funnel. That is a typical marketplace pattern: users browse inventory first and commit later.

The Admiral Media Liquidity Loop Framework

The Admiral Media Liquidity Loop Framework is a seven-step method for planning paid acquisition in two-sided apps. It is built around one idea: every euro spent on one side should increase the chance that a user on the other side finds a match. Admiral Media uses it to sequence channels, events and creative so that growth on one side does not outrun the other.

1. Map the sides and their intents. List every user group the app serves (for example renters, buyers, agents, owners) and the job each one hires the app for. Treat each intent as its own audience with its own message, keywords and events.
2. Find the constraint by segment. Check supply depth, time to first match and repeat behaviour by city, category and price band. Budget goes where demand can actually be served.
3. Define the liquidity event. Pick the first interaction with supply that predicts value, such as an enquiry, a saved search, a lead or a first order. This becomes the north-star event for paid media.
4. Build an event ladder. Start optimizing to an event with enough volume, then step deeper as data accumulates. Never jump straight from installs to the rarest event.
5. Match the store page to the intent. Send each intent to a store page and in-app destination that reflects what it searched for, using custom product pages on iOS and custom store listings on Google Play.
6. Run creative per side, not per channel. Test messages built around each audience’s motivation (price, choice, speed, trust) and let the winners travel across channels.
7. Close the loop with retention and re-engagement. Bring back users who searched but did not transact, and measure incremental lift so retargeting budget does not simply claim credit for organic behaviour.

The rest of this guide walks through how each step works in practice, with results from Admiral Media’s marketplace and listings clients.

How did ImmoScout24 cut CPA with custom product pages?

How do you market one app to renters and buyers at the same time? You give each group its own store page, matched to the keywords that reveal its intent. Admiral Media helped ImmoScout24 build an Apple Custom Product Page strategy for Apple Search Ads, and the custom pages reduced cost per acquisition for both renters and buyers compared with the default page.

ImmoScout24 describes itself as the number one real estate app and digital real estate marketplace in Germany. According to the ImmoScout24 case study, the brand has 98% brand recognition and more than 20 million users each month across web and app. The challenge was a classic marketplace one: two user groups, renters and buyers, with different intentions, customer journeys and behaviour, and an App Store that previously allowed only one rigid product page.

Apple introduced custom product pages with iOS 15, and they integrate directly with Apple Search Ads. Admiral Media worked with ImmoScout24 to define the audience groups and their intent, helped create the pages, and tested them against the most relevant search terms. When a user searched for “wohnung mieten” (rent an apartment), the ad led to a page that matched that search, instead of a generic page that tried to serve every visitor.

The results below come from relevant ad groups in Generic campaigns over four weeks in August 2022:

• Buyers: CPA from €2.41 to €1.36: the custom page outperformed the default variant for people looking to buy, a reduction of about 43%.
• Buyers: CPT from €1.93 to €1.13: cost per tap fell by about 41% for the same ad groups.
• Renters: CPA of €0.80 versus €2.15 for the default page in the same timeframe.
• Renters: CPT from €2.00 to €0.69, while tap-through rate rose to 28% versus 22% for the default page.

Why does this work? In Apple’s search auction, relevance affects how often an ad is shown and tapped. A page whose screenshots and copy match the searcher’s words earns more taps per impression, and a higher tap-through rate lowers the cost of each tap. Then, after the tap, the user sees the exact promise they searched for, so more taps turn into installs. The renter result shows both effects at once: CPT fell and TTR rose.

The feature has grown since this campaign. The case study describes up to 35 custom pages at launch. Apple’s current custom product pages documentation says developers can publish up to 70 additional versions of their product page, assign keywords so a custom page can appear in organic search results for those terms, and attach a deep link to each page on iOS 18 and later. For a marketplace, that means each intent can now get its own store page, its own search visibility and its own landing point inside the app. Admiral Media explains why these pages should not be treated as A/B tests in the article why Apple custom product pages are not A/B tests.

Alexandra Zinnow, Lead PPC, Digital Marketing at ImmoScout24, summarised the engagement this way: “Admiral Media were extremely helpful and efficient in making Apple’s Custom Product Pages work for us. As a result, we were able to decrease our CPI and improve the overall performance of ASA campaigns.”

How should a marketplace app structure its store pages and landing points?

What should a marketplace do with its store presence? Treat the store page as part of the funnel for each audience, not as a single brochure. Every major intent should land on a page that names its need, shows the relevant inventory and opens the app at the right screen.

A single default page has to talk to renters, buyers and agents at once, so it says something generic to all of them. In the accounts Admiral Media manages, the most productive structure usually follows search intent rather than demographics:

• One page per core intent. Rent versus buy, sell versus shop, compare versus manage. These are the splits users express in their search terms.
• Category or location pages where inventory is deep. A page for a specific property type, product category or city only makes sense where supply can back up the promise.
• Supply-side pages. If the app also recruits agents, sellers or providers, they need their own page. A seller does not want screenshots of a buyer’s search results.
• Deep-linked landing screens. The page and the first in-app screen should match. A renter who tapped a “flats for rent” ad should open on a rental search, not on a home screen that asks them to choose again.

Google Play offers a comparable tool. According to the Google Play Console documentation on custom store listings, developers can create custom listings for specific audiences and target them in several ways, including by country and by Google Ads campaign. The principle is the same on both stores: the more closely the store page matches the user’s reason for searching, the less work the ad has to do.

Which paid channels work best for marketplace apps?

Which channels should a marketplace app use? The channel mix should follow intent. Search-led channels such as Apple Ads and Google App campaigns capture people who already know what they want, while social channels such as Meta and TikTok create demand and test which motivations move each side of the market.

Many marketplace users arrive with a clear, searchable need (“rent apartment Berlin”, “compare car insurance”), while many future users have not started looking yet. A balanced plan serves both. The table below shows how Admiral Media typically assigns roles to channels in a marketplace account. It describes roles and decision criteria, not performance benchmarks.

For Android-heavy markets, Google App campaigns are usually the backbone. Admiral Media’s Google App campaigns guide covers how bidding moves from installs to in-app actions. Re-engagement is especially valuable for marketplaces, because so many users search, leave and come back when their need becomes urgent. Google states in its announcement of App campaigns for engagement that these campaigns are built to stay connected with existing users and take them to relevant content in the app, with eligibility tied to a minimum base of 250,000 installs at launch. Larger marketplaces almost always qualify.

How did kaufDA grow user activity with creator content on TikTok?

Can creator content work for a marketplace? Yes, when the creators show real inventory in use. Admiral Media ran TikTok campaigns for kaufDA, a German deals app, and found that influencer creatives outperformed regular ads on user growth, user activity and cost per install.

kaufDA lets users browse shopping deals on groceries, cosmetics, furniture and electronics, follow favourite stores and set alerts. In marketplace terms, the supply is the offers retailers publish, and the demand is shoppers looking for savings. The goal, according to the kaufDA case study, was to grow app downloads and the user base by finding the best and fastest scaling acquisition channel.

After initial tests with in-house creatives, Admiral Media acquired content from TikTok creators and identified two winning approaches: inspiring, authentic, good quality content for shopping lovers, and targeting an audience that matched the persona of the typical kaufDA user. Within one month the campaigns generated almost 70 million impressions in Germany across Android and iOS. Comparing regular ads with creator ads, the case study reports:

• +1000% user growth on one specific day, driven by creator content.
• +731% user growth overall for creator ads versus regular ads.
• +146% growth in user activity for creator ads versus regular ads.
• -18% CPI, with a higher engagement rate per impression.

The lesson for marketplaces goes beyond TikTok. Inventory is the product, so the best creative usually shows inventory in a person’s hands: a creator flicking through this week’s offers, walking through a listing, unpacking an order. Admiral Media’s guide to UGC ads for mobile apps covers how to brief and test creator-style content so it converts rather than just entertains.

How do you build creative for two sides of the same marketplace?

How should creative differ by side? Each side needs creative built around its own motivation, because a renter, a buyer and a seller are persuaded by different things. Admiral Media briefs creative per audience and per motivation, then lets the winning concepts travel across channels.

Demand-side users usually respond to choice, price and speed. Supply-side users respond to reach, revenue and simplicity. Trust runs through both: a buyer wants real listings, a seller wants genuine enquiries.

Admiral Media’s work with Clark shows how wide the creative range can be in a trust-heavy category. According to the Clark case study, the team tested tangible benefits (price, no paperwork, number of existing users) against intangible ones (feeling safe, trustworthy expertise, less complexity), and tried selling points from simple bike insurance to house insurance. The team quickly identified the big winners and doubled down on them.

Admiral Media’s creative testing framework explains how to structure those tests so winners are scaled before they fatigue.

How did Clark lower cost per lead by 50% in a trust-heavy category?

How do you grow leads when users hesitate to share personal data? Combine broad creative testing with careful audience broadening and smarter optimization events. Admiral Media used this approach on Facebook Ads for Clark, a German insurance management app, and reduced cost per lead by 50% between month one and month three.

Clark was founded in 2015 to move insurance into the digital era. The app lets users manage and compare their insurance policies, review contract details and get tips on improving their coverage. When Admiral Media started, install volume was healthy and leads were stable, but lead growth needed to improve. The strategy focused on Germany and Facebook Ads, with three components described in the case study:

• Diversified creative with rapid testing. Many creative routes were tested, and winners were scaled fast.
• Controlled audience broadening. Combining genders, age groups, interests and lookalikes increased leads among Android users, women and younger consumers, audiences that had been under-indexed before.
• Smarter optimization events. The team optimized toward events on the side of the main funnel, and paid special attention to the middle funnel step “level achieved”, which is a common drop-off point because people are sensitive about giving out their phone number.

Comparing month three with month one, Admiral Media’s work with Clark achieved:

• -50% cost per lead, the main business goal.
• +41% conversion rate from the improved creative and audiences.
• -29% CPI and +18% installs.
• +13% count of level achieved and -47% cost per level achieved at the sensitive middle funnel step.

The case study also notes that month two (November 2020) worked slightly less effectively because of Black Friday, and that a quick optimization early in month three produced the best install to lead conversion ratio. Plan creative and bids for shopping-season auction pressure rather than reading a seasonal dip as a structural problem.

How do you scale a transactional marketplace from zero to large budgets?

How do you scale spend without wrecking efficiency? Build a creative production engine first, then climb the event ladder from installs to sign-ups to first orders as the accounts gather data. Admiral Media used this approach for a UK fast commerce app and took it from zero to seven figures in monthly user acquisition spend.

The client was a leading grocery delivery company in the UK, operating in neighbourhoods across London, Amsterdam and Paris, offering 24/7 ordering, more than 1,000 essential items and delivery in under 20 minutes. Quick commerce is not a classic two-sided marketplace, because the company holds its own stock. On the demand side, though, it behaves exactly like one: value depends on local availability, and the first order is the event that proves the model to each user. According to the fast commerce case study, Admiral Media assembled a team of marketers, designers and data analysts and built three things.

• A creative production framework. The team brainstormed ideas and analysed the market’s creatives twice a week, produced dozens of storyboards, then built, resized and launched tests using an accelerated testing strategy designed to reach statistical significance within one or two days.
• A broader creative analysis. Winners were judged on more than CPA, using CPM, CPC, CTR and install conversion rate where relevant, and then scaled with variations.
• A funnel approach to optimization. Campaigns started with install optimization, moved down to sign-up optimization and later to first order optimization once the accounts had enough learnings and data. The team then added new ad channels to diversify the acquisition stack.

Admiral Media’s work with the fast commerce app delivered these reported results:

• +64% paid installs: the smallest gain, because installs were never the goal.
• +5152% paid users and +2178% paid conversions as optimization moved down the funnel.
• +1210% paid purchases, the transaction event that proves value.
• -91% paid CAC over the same engagement.

The gap between install growth and purchase growth is the point. Moving optimization from installs to sign-ups to first orders taught each ad platform who the valuable users were. Healthy marketplace scaling has this shape: deep events growing faster than shallow ones. Admiral Media’s guide to scaling app user acquisition covers budget increases and learning phases in more depth.

How should marketplace apps measure success beyond installs?

What should a marketplace report on? Report on matched outcomes by side and segment: cost per liquidity event, time to first match, repeat rate and the share of new users who transact. Installs and CPI are inputs, not results.

A common reporting mistake in marketplace accounts is a blended dashboard: one CPI, one ROAS, one CAC across every audience. A renter and a buyer, a shopper and a seller have different values and costs, and averages hide exactly the differences that drive the business.

A useful measurement stack for a marketplace app has four layers:

1. Platform signals for optimization: the events each ad platform receives, chosen for volume and predictive power.
2. Attribution through a mobile measurement partner for cross-channel comparison, with SKAdNetwork or AdAttributionKit conversion values on iOS mapped to the liquidity event rather than to installs. Admiral Media’s guide to SKAdNetwork conversion values explains how to design that mapping.
3. Cohort economics by side and segment: how much revenue or take rate each acquisition cohort produces over time.
4. Incrementality for retargeting and brand search, where the risk of paying for users who would have come anyway is highest. Admiral Media’s guide to incrementality testing covers how to run holdout tests.

What are the most common marketplace app marketing mistakes?

Where do marketplace apps usually waste budget? Most waste comes from treating a multi-sided app as if it had one audience and one funnel. The same handful of mistakes appear again and again in the account audits Admiral Media runs for marketplaces.

• Buying demand where supply is thin. Scaling campaigns in cities or categories with too little inventory produces expensive users who leave after one empty search.
• One store page for every intent. A generic page makes renters, buyers and sellers all work harder to see themselves in the app. ImmoScout24’s results show what matching the page to the intent can do.
• Optimizing to installs for too long. Installs are easy to buy and weakly tied to value. The fast commerce results show how much faster deep events grow once optimization moves down the funnel.
• Creative that shows the app instead of the inventory. Users buy access to listings, offers and providers. Screens of empty interfaces rarely persuade.
• Retargeting without incrementality tests. Marketplace users return on their own when a need comes back. Without holdouts, retargeting can look far better than it is.

Admiral Media’s mobile app marketing agency team works across paid media, App Store Optimization and creative production, because the biggest marketplace gains often sit between those disciplines. ImmoScout24’s results came from a store-page decision executed inside a paid search channel, kaufDA’s from creative sourcing, and Clark’s from creative, audiences and event design working together.

The learning that travels across these accounts: a marketplace grows when each side finds what it came for quickly. Paid media brings people to the door. Whether they stay depends on what they find inside.

Frequently Asked Questions

What is marketplace app marketing?

Marketplace app marketing is the work of acquiring and retaining both sides of a two-sided app, such as buyers and sellers, renters and landlords, or shoppers and retailers. It differs from single-sided app marketing because each side’s experience depends on the other side being present. Campaigns are usually optimized toward the first meaningful interaction with supply, such as an enquiry, a saved search or a first order, rather than the install. Store pages and creative are built per audience, because each side responds to different motivations.

Should a new marketplace app spend on supply or demand first?

Spend should go to whichever side is the constraint in each segment. A young marketplace with plenty of interested buyers but few providers should fix supply before buying more demand, because demand without supply produces empty searches and fast churn. An established marketplace with deep inventory is usually limited by demand activation, so paid media goes mainly to demand. Checking search to result rates and time to first match by city or category shows where the constraint sits.

What is the best KPI for a marketplace app’s paid campaigns?

The best primary KPI is usually the cost per liquidity event, meaning the cost to get a new user to their first meaningful interaction with supply. Depending on the model, that could be an enquiry sent, a lead submitted, a store followed or a first order completed. CPI is useful as a diagnostic, but it is weakly linked to marketplace value. Over time, cohort revenue or take rate per acquired user should become the scaling metric.

How do custom product pages help marketplace apps?

Custom product pages let a marketplace show a different App Store page to each audience, so renters, buyers and sellers each see screenshots and copy that match their intent. Apple says developers can publish up to 70 additional product page versions, assign keywords to them and attach deep links on iOS 18 and later. In Admiral Media’s work with ImmoScout24 on Apple Search Ads, custom pages lowered buyer CPA from €2.41 to €1.36 and renter CPA to €0.80 versus €2.15 for the default page over four weeks in August 2022.

Does influencer or creator content work for marketplace apps?

Creator content tends to work well for marketplaces when it shows real inventory in use, such as a creator browsing current offers or walking through a listing. In Admiral Media’s TikTok work for the German deals app kaufDA, creator ads delivered +731% user growth overall, +146% user activity growth and an 18% lower CPI compared with regular ads, with almost 70 million impressions in one month. The key is briefing creators around the audience’s motivation and testing several approaches before scaling.

How long does it take to see results from marketplace app marketing?

Timelines depend on tracking quality, inventory depth and how much data the accounts can generate. In Admiral Media’s work with Clark, the comparison of month three with month one showed a 50% lower cost per lead on Facebook Ads, after a weaker second month affected by Black Friday. Structural fixes such as intent-specific store pages can show effects within weeks, as ImmoScout24’s four-week test did.