We spent $20k on Google Play pre-registration ads | Hacker News Reader

We spent $20k on Google Play pre-registration ads

andreaskambanis.com

207 pointsandreaskam131 comments

> MealPrepPro is built by a small indie dev team and $20k was a painful amount of money for us to waste. With pre-registration ads you're putting all your faith in Google that they'll deliver the conversions they're showing in the ads reports

I am running a small independent startup myself where I count every nickel and dime. I tried Google Ads (there was an offer to spend $400 to get an additional $400). Besides a few sign ups, it didn't pay off at all. I would never even consider spending $20,000 (!?). I talked with a few newsletters (starts at $1500 per quarter for a campaign) and social influencers (starts at $700 per post) and if money wasn't an issue I would buy marketing, but in the early stage the most important resource is myself. If I spend my savings that I live off, I risk running out of money and must go back to being a freelance developer (and the European market sucks at the moment).

Their biggest problem was not doing research. Conversions are hard, usually a percent or two. But what is a conversion? It's getting a user to respond to your CTA. In this case, the CTA was to get people to see the ad and click through to a "preorder" page. They got what they bought. The problems were they expected a greater response, didn't know what they were getting into, and spend $$$ on ads before anyone could get the product. That was the real killer, that's why the money was wasted. They thought they were spending 20k for installs, they were actually spending 20k on were clicks to an app no one could get.

Don't spend money on preorders if you're not collecting payment data now and authorization to charge it later. Anything else isn't a preorder, it's a waste of time.

> Anything else isn't a preorder, it's a waste of time.

cept for tesla, they charged $100 for a "pre-order" (of the cybertruck).

It's smart, because they managed to get an interest free loan off these customers, and also gauged interest at the same time!

So yes, the CTA must include a form of payment - otherwise, you're not getting a customer for your spend, you're just getting eyeballs.

Just thinking out loud - is there any way to create a profitable Google Ad campaign without sinking thousands of money ramping up and learning?

It doesn't appear so.

The best way to learn is to become a PPC expert at some marketing company, you'll still burn through thousands learning all the tricks, but at least they won't be your thousands.

I haven't fully validated this idea yet, but I ran a Google ad campaign for $5 a day for a mailing list signup page that I was planning a business around. Click through rates were around 2.5%, which seems to be the baseline, i.e. a 2.5% click through rate basically means no engagement. One thing that was interesting, however, is that certain keywords had click through rates around 7%. I wanted to try re-targeting the campaign to just those keywords but I lost motivation.

Was this a long time ago, or for a totally empty niche? I ask, because I'm lucky if I can buy a single click through for $5, and it's been like that for several years now.

This was a few months ago. I don't think it was totally niche. It was about $1 per click through and I run the campaign for 2 months, which was an accident. I meant to pause a lot sooner.

Generally yes, it's just a different approach. Most people in the Google Ads space actually get their start with warm intros to companies and/or individuals that want to get into digital marketing, and just get a percentage of their monthly ad-spend for contribution. You can haggle with price per lead or revenue share from the leads that convert, but you still need capital to initially launch those campaigns.

When I started out in lead-generation using Google Ads for ACA coverage, Final Expense insurance, roofing and solar installs, I did it in this order:

1.) Reached out to insurance agencies & roofing companies I already knew, convinced them I could do their digital marketing campaigns for a percentage of their monthly ad spend;

2.) Got upfront capital through that ad spend percentage and had the clients manage the ad spend for their own campaigns, and used the ad spend percentage capital to build cost per-lead campaigns for other clients. These were far more lucrative in investment, but cost a lot of money upfront.

3.) Proved my merit for clients in both approaches, and used my second quarter audit to "upgrade" the clients in number 1 to a cost per-lead, and negotiated with clients in number 2 to a revenue share for converted leads.

It's a grind, it's hard, but it can be done.

"is there any way to create a profitable Google Ad campaign without sinking thousands of money ramping up and learning?"

Yes, it's called hiring an expert with a track record who will get you profitable a lot faster than you probably could yourself (assuming you are inexperienced).

Where do you find these experts? I've tried managing campaigns myself and concluded the same thing you did, but have struggled to find an expert. There are so many people online who claim they can run your campaign but I have a hard time separating the real experts from the "fake it til you make it" types.

It's difficult to find, especially when you are working with a limited ad spend. Why would someone skilled run a 30K month campaign for you, when they could make a lot more money running a 300K per month campaign for someone else.

When working with smaller numbers, you are going to be working with someone who is a little less experienced, more than likely. So, finding someone who can share a solid, provable track record may be a little harder. Figuring out a compensation scheme that takes their inexperience into account and shifts some of the risk to them as a result of this is key. You cannot enter a deal in which they just take a percentage of ad spend, because if they don't produce sales, then they get their money and you lose big.

I guess what I am trying to say is that you need to focus on the actual "deal" that you make with the person running your CPA campaigns to ensure that the risk is a bit more balanced and incentives are aligned...

It’s very hard to find good talent under $5k/mo. At those rates and above usually finding a referral works.

UpWork is spotty.

How do you prove out their track record?

They can work well for B2C, wouldn't expect the same for B2B businesses.

I think it is possible, but you can't do it on the freemium model, or very aggressive paywalls for premium features.

I have spent a few $00 possible $000 on Google Ads and had nothing to show from it. Lots of traffic no sign ups. I then emailed a list and got a 4% conversion, so it isnt the product, it must be the targetting. As Bob Hoffman (a contrarian ad thinker) says do you want to die wasteful (spend $ on analogue advertising) or unknown (spend $ on line advertising)

It seems pretty clear that's what you get when you use the campaign? https://support.google.com/googleplay/android-developer/answ...

"After you make an app or game available for pre-registration, users can visit your store listing to learn about and pre-register for your new app or game.

Then, when you publish your app or game later, all pre-registered users will receive a push notification from Google Play to install it.

Eligible devices will also have the app or game auto installed on the day it launches. (more details on this in documentation)

I don't know anything about app releases, and I'm sure something like this is important to get an app into trending lists.

However I'm a bit skeptical that auto-downloads would even really be better for a free app. If the user isn't willing to accept a notification to install an app, they probably aren't going to use the app if it just auto-downloaded.

Also this just seems like a very weird app to have pre-registration ads for. If a user is looking for a "meal planner" app they obviously aren't going to just wait 2.5 months for your app (which doesn't seem particularly unique) to come out. They're going to download and try other apps that are actually available during that time.

It still seems like a rip-off though.

If you look at the first screenshot in the article it shows what the pre-registration page looks like.

It says:

Perks of pre-registering

- Automatic install

Install automatically when it's available

>If you choose to make your app eligible for auto install, Play can deliver your app to users’ devices automatically on launch day (if they’ve opted in).

It's up to the users.

And you need to still do quite a lot of leg work getting users to opt in. With the recent death of the Reddit API my favored client is being adapted to Lemmy - but the news of that happening only reached me because I was following the news on their subreddit before the shutoff. Pre-registration feels like a huge waste unless your users are literally chomping at the bit.

Users install a whole bunch of apps on their phone when looking for a solution to a problem and a fair few of those are never actually launched.

"Would you like to install this meal prep app in a couple of months time?" is a question I cannot imagine anyone saying yes to. Why would you? There are apps available today...

Is sending someone a push notification really worth $1 though?

Is paying $1-3 simply for a click on google ads worth it either?

(In my experience, no).

I don’t have experience but I thought Subprime Attention Crisis was an interesting read

Depends on your conversion rate and customer LTV.

Check Mesothelioma ad click rates on Google...

Yep, and what your CPA is.

LTV for a 30 yr mortgage is hundreds of thousands - that pushes for some expensive clicks at auction.

To the users who have actively indicated that they are interested in downloading your app, all at the same time on launch day, potentially putting you on some "most downloaded apps" leaderboards? Quite possibly, yes. It'd depend a lot on the app, though.

Neat article.

I would like to criticize your landing page for your app. It very very much bugs me, when landing pages for apps have only a QR code which I am supposed to scan with my phone.

Luckily, you also added a "Get the App" button. Problem is, that you try to be smart and automatically forward me to either the Google Play Store or Apple App Store. Since I opened the link on a Macbook Pro, you assume I want to see the App Store. However, I am an Android user.

I just want to be able to see two buttons, one for each store and click it myself.

You are by far not the only one who does this. Would be great if App devs would change it all together.

Yeah. The landing page didn't answer the questions I had. It's a hard pass. Feedback:

- Give a long enough free trial to get addicted (e.g. 30-90 days). If you're delivering value, I will pay $4/month

- For me to have value, you'd need to plan things like ingredients. In a best-case scenario, you'd integrate with Instacart and other delivery services, and there would be zero waste. You'd also make use of ingredients I already have.

I have no idea what it is you actually do, though, so I won't install. I can already subscribe to a recipe listserv, and I have no idea the delta. Is this the same thing, only for $$$ and with aggressive data scraping from my mobile?

You don't need to answer here. Answer on your landing page. I was curious around to click around, and find crickets.

Thank you for posting and sharing, though.

Yeah we often think about the free trial vs freemium experience. I really admire apps like FitBod who just let you try out 3 free workouts without needing to commit to a free trial.

Eating healthy is one of the toughest habits to commit to. The benefit of the 7 day free trial is it really pushes people to try it out during those 7 days and see if it's for them. Otherwise you download and think "I'll look at this later" and then never come back to it.

Not saying what we have is perfect and I'm still looking for better options. We're always down to give people a longer trial if they get in touch with us. I get that 7 days won't be enough for everyone.

RE: Instacart integration - we already integrate!

> RE: Instacart integration - we already integrate!

From the point of view of a potential user deciding whether to convert, no you don't!

Do a few cafe studies where you buy random people a coffee in return for exploring your landing page and figuring out (1) what you do (2) how it would integrate with their life (3) what they need.

I'd like a product which takes thinking out of eating cheap and healthy, and if you do that, your app will pay for itself. I have no idea if it does.

This is definitely something I was cognizant of when I made my product page.

Since I have an app for both platforms, 2 buttons works very well, allows the user to make the choice, and you can even add a simple useragent check and redirect the user after X seconds if they haven't actioned.

There are some interesting "rules" (or recommendations) around how to place app store images, their sizes, etc, too.

I'm using Brave on a Windows PC, the "Get the App" button sends me to the Apple store also. In fact, there is no indication that this is even available on Android at all.

How many people out there install smartphone apps from a desktop web browser? You anyways have to pull out the phone to use the app. Having a QR code front and center and navigating to the app store listing on the phone is the preferable outcome for 99% of customers. The majority of them aren't even signed in to Apple/Google on the desktop browser.

I almost exclusively install apps from my desktop browser. Is that not normal? I find it much easier to verify that I'm looking at the app that I intend.

I do. But even if I didn't I might just want to see the play store page without jumping through hoops.

Most people are already on their phone when they see your website. Are they supposed to pull out another phone to scan the qr code?

When they're on their phone it'll show a download button instead :-)

The QR code only appears on desktop.

Again, what do you gain with inconveniencing the user and adding friction, instead of just taking them to Play Store and letting them install with one click?

You can take a screenshot and long click the QR to go to the link. At least on iOS. No need for another phone ;)

Not exactly obvious to say the least, but it works.

There's already a download link next to it.

One point that people without app experience probably aren't considering - on iOS, having a large number of downloads (particularly in the first week after launch) can really boost your keyword rankings and ASO. Without this boost, climbing up the keyword rankings can be extremely difficult.

Apple's pre-order feature does precisely what I'd expect - automatically downloading the app when it's available, which helps jumpstart the flywheel explained above.

I would have expected Google's pre-registration feature to work similarly. Thanks to OP for spending the $$ to find out this isn't the case!

haha it's been my pleasure being a guinea pig. Just wish I could of learned the same lesson for $5k instead of $20k.

One question I had when reading: Had you opted in to all of the appropriate settings so that your personal device should have automatically installed?

I have no way of being 100% certain, however I followed the best practises: - Being on a new version of Android - Having an up to date version of Google Play store app - Being on wifi - Having lots of battery and also plugging my phone in to power

Short version: We used a Google Ads pre-registration campaign to get installs for our app. Google charged us for 16,171 conversions. When we launched, we only received 1,371 installs.

The core reason seems to be that Android pre-registration (unlike iOS) doesn't, by default, automatically install the app once it's available. Instead it just signs the user up for a push notification, unless the user has opted in to automatic installs. So this campaign had 16,171 people click through and pre-register, but only 1,371 of those converted to actual app installs once it was available.

And we threw $20k in ad spend at it (which quite frankly is insane).

Seems like at this point you'd come out ahead if you just mailed people 2 dollar bills with the promise of another 2 dollar bill if they install your app

Honestly, there's a lot of advertising out there that I think could benefit from direct monetary incentives. A large portion of the population is becoming resistant to regular advertising (good for them) and opting in to ad blockers when available. Actually paying people for the attention you're demanding from them would be refreshingly different... and heck, it worked for timeshares.

That's a good point. I got some in-game spam mail in World of Warcraft the other day with a few gold coins attached. Even though it was an amount you'd earn in about 2 seconds of playing the game (real world value: 4/100ths of a cent), it made me a lot less angry than ads usually do

Indeed there are groups and "businesses" that "pay" you to install apps. I assume someone is on the other side of that transaction paying for new user installs. Doing it yourself would probably be cheaper and you would probably be offering a much higher payout than those systems.

It's called "incent traffic" and it's notoriously awful traffic. Users are worth 5% of a normal user because they're freebie seekers who are usually broke.

What was the time between pre-registration and release?

We started the pre-registration around 2 months before release and it ran right up until release. The max you can do is 90 days.

This doesn’t surprise me. A user has a need - in this case a food diet app. They find this app, think it looks good, and registers their interest. Good.

Launch day comes around, they’re nowhere to be seen - why? Well, they may have found another app, or decided they don’t need the app after all - say they lost interest in the diet / health kick.

I feel like app registrations would only work for “special” apps that don’t have close competitors or have a particular buzz - e.g. Threads, ChatGPT etc

EDIT: I’ve misunderstood - the app was supposed to auto-download on launch day. It didn’t. This story makes a lot more sense now! I assumed it just sent a push notification / email.

I work on related things at Google so have to be careful what I say here, but to clarify things a little...

In the docs it makes it clear that there are a bunch of caveats to the auto installation. Reading through these caveats, they are all there for good UX reasons, essentially to ensure that what gets advertised is what gets installed (no bait-and-switch), and that users are going to be ok with the auto-installation (doesn't eat all their data or battery life). There is advice on how to improve eligibility, it's not clear if the author followed this or not.

The eligibility factors will vary from user to user, and may between user bases as well. For example if you have an app that is popular with users on devices with less storage space, there are likely to be more who can't install it because they don't have enough space.

One or more of these factors may not have been met for the author/publisher of this app, and it's also possible that their target market also doesn't meet the conditions for auto-install as much.

Appreciate the reply.

I think you're referring to these guidelines: https://support.google.com/googleplay/android-developer/answ...

We did go through the guidelines to make sure our app would be eligible for auto install. For example our download size is sub 100mb. I can see in Google Play Console that all our installs are Android M+.

There are some factors I can't verify. For example, where can I check what Google Play Store version users are on?

If certain factors make a device ineligible for auto install, then maybe these devices should be excluded from the ad campaign to help ensure developers are getting good value for their ad dollars?

It created such a negative first experience for us on the Google Play store, which is a shame as we'd of been happy to invest in install ads and Google would of made a larger return from us over time. The low quality of support gave us no path to at least share our concerns with Google.

Alternative idea: spend $1,000 or less on something that is unproven. You speculated and it turned out different than expectations. That's the name of the game. The only right answer is test. But you shouldn't expose yourself to so much downside.

App launch is a unique situation: you can only do it once. As such, pre-launch advertising is always unproven. You would have to have high confidence in the integrity of the ad network.

Since online advertising is essentially untrusted, with an entire industry dedicated to auditing its claims, I think the correct approach is to never try this form of unprovable advertising.

Was this really an option here? They probably aren't launching that many apps on the Play Store. So they had one shot to try out a pre-registration campaign to get a big app launch.

For sure. Live and learn. And hoping to share it here so others can learn from my mistake.

This sort of ad just doesn't seem useful for this type of app.

I'd think most people would want a meal planning app that they can use right away, so they'd just download a different app when they see yours isn't available yet.

I wonder how many of those "pre-registrations" thought they could install it right away, saw they couldn't and then just found something else.

All good points. Especially about meal planning being something you're looking to solve right now, not in a few weeks time. Unlike a game, where a few weeks from now you'll still be looking for new entertainment options. Thanks for adding your thoughts, I think they'll be helpful for someone else considering pre-registration for their app.

Conversion rate issues aside, why did they choose pre-launch ads over post launch advertising?

Some apps rely on a critical mass of users on launch (multiplayer games, social networks, dating sites, etc.) so generating hype pre-launch is important for these.

But this looks like an app where it doesn't matter for a user how many other users there are. So slowly ramping up your ads after launch, while you work out bugs and other flaws looks like a much better approach to me. This also allows you to track ad effectiveness and adjust as you go, instead of taking a big gamble.

We felt that getting a lot of traction on the launch day might give us some good momentum in the charts. With hindsight, I'd wish we'd saved our money and spent it after launch, when we could sense check the figures that Google Ads is reporting.

Or target a few low competition keywords with custom landing pages and get the traffic for "free". Looks like Google has plenty of traffic in that segment and low competition.

Not just google ads. Meta ads and reddit ads all fake the metrics and charge more money than what they send you. It's all a scam but because of the larger gun policy nobody can do anything.

Google has such a gigantic incentive to lie about how effective their advertising is. It's impossible to verify and they make more money the more effective they say it is. So it's something I've suspected is happening for a long time.

The question then becomes, what does the internet look like if advertisers only spent 10% of what they do now? How much of the internet is built on the lie that digital advertising works? What services would have to shut down? How valuable is user data really? If user data isn't valuable, because advertising isn't valuable, then what else on the internet becomes impossible financially?

Many, many websites function based on the assumption that user data and advertising will always be money makers. What if they aren't? What if we've all been lied to about the effectiveness of ads?

In 1995 the Internet existed, it was basically great, and basically no website was ad funded.

I cannot think of a single essential service I use on the Internet today which is ad funded. If it's essential I'm paying for it because the free, ad-funded version is shitty and unreliable.

We could ban Internet advertising tomorrow (give people time to migrate to paid services) and I genuinely don't think it would have much negative impact on society. Social media like Twitter and Meta would take a big hit and that would be very good for our society. Less teen girls would commit suicide, and less hate would be spread.

I'd start paying ten bucks a month or whatever for Google (already pay them a pretty penny for Workspace) and can't think of much else that I'd lose that would matter. Subscriptions to a couple sites of professional interest like Stack Overflow? Small price to pay for ending Internet ads.

I'd also add that personally motivated hosting is basically what gets small sites through the day. I worked on a MUD for quite a while where a small portion of the playerbase was working professionals and the majority of the playerbase was broke college students with far too much free time. The working folks would shell out a completely inconsequential 30 dollars a month between ten of them just so they could have fun playing the game - there was never any serious discussion of adding a paid requirement to join because 3$/month is very little for a steady stream of entertainment and because the "leeches" actually contributed immensely to the activity in the MUD. The workings folks contributed money, the college folks contributed time and everyone had fun.

If you're an independent reporter running a blog you can easily cover the hosting costs by just writing an occasional article for a major organization. If you have a fun hobby that hobby is probably worth the cost of hosting a server to attract other hobbiests.

Really, the only thing that would possibly die forever might be services like YouTube which have absolutely absurdly weak monetization potentials when compared to their infrastructure costs. Losing YouTube would suck, but if it meant I'd never need to see another internet ad I think the cost is worth it... and more curated services like Nebula have proven that purely subscriber funded video content can work - but it'd be hard to enter that market with no free hosting platform like YouTube.

> If you're an independent reporter running a blog you can easily cover the hosting costs by just writing an occasional article for a major organization.

I agree with most of what you wrote except for the above. Unless you're an independent reporter in things like gardening or motor sports, the major media organizations see you as an enemy and would prefer to shut you down.

> In 1995 the Internet existed, it was basically great, and basically no website was ad funded.

Sorry, but which parallel universe is this?

Advertising was definitely less intrusive, but I would say that an even larger proportion of websites was ad-funded than today.

A shitton of content is (and was) on platforms which are entirely ad-funded. Think geocities, the various blogging platforms, and twitter et al. today. You remove ads, you force the authors of this content to pay for their publication (a ridiculously cheap amount sure, but the large majority of them won't pay a dime).

I have paid for hosting my personal/hobby websites since approx 1993, but I know very well I'm a 1 to 10000 exception, based on the ratio of people who participate on my website.

Ads were of course common in 1995, but monetisation in general was rarer. A greater proportion of the web was hobbyists writing and collating information about subjects they were interested in, not because of financial incentives.

I mean, isn't that basically what the first .com was about? :)

Everyone went from thinking every page view was equivalent to a magazine/tv/radio ad view, paying 5-10$ CPM and then everyone wised up and the industry was decimated for a while.

The first .com bust was that these e-stores weren't profitable compared to valuations. So much money was poured in.. I use to get 1 or 2 dollar checks from free spin places. Too much money too soon. Ads had nothing to do with the bust. When pets.com went bust the market the air left with it and everyone moved into companies laying cable.

>It's impossible to verify

1. Determine sales rate 2. Deploy ads 3. Determine delta sales

Am I missing something?

Confounding effects

Wouldn't that be a good reason not to spend money on ads?

>What services would have to shut down? How valuable is user data really? If user data isn't valuable, because advertising isn't valuable, then what else on the internet becomes impossible financially?

If it suddenly stops being the case that everything is judged against an ad-supported but free to use baseline, I think paid services become more viable than they currently are. There would probably need to be some work on payment models, but I think not many things would be forced to shut down (ignoring whatever shutters while the dust settles).

> It's impossible to verify

Spend on ads is extremely easy to verify. For every ad channel you use a different discount code. Then you see how many customers used a specific discount code when making a purchase. Discount codes work everywhere: Social media, search ads, radio, TV, banners, etc etc

If you don't use discount codes when advertising it means you don't care for how you spend your marketing money - which is the truth for a large majority of businesses. They simply don't care if their advertising works or not.

That was the past, which gave us a glorious internet at the expense of idiotic advertisers in exchange for a few bits of data.

Thanks to the EU and the privacy social justice warrior that era is gone: welcome to the era of cookie banners and walled gardens.

Search will be even worse than today, your Google search won't be very useful (and it's definitely not as useful as it was 15 years ago): you'll have to search on Reddit, Facebook and who knows how many other walled gardens which can't live off advertising anymore and will need to find a different revenue model (which probably will cost the user).

Hang on... the decline in Google search is due to the EU?

Walled Gardens are caused by cookie banners?

I have heard some things in my time....

You just have to run your own measurements and not rely on their data.

> because of the larger gun policy nobody can do anything

Oh please do elaborate

You'll spend more than $20K trying to get your money back in court.

Ah so it’s a joke used to reference the complete lack of power companies/people have against behemoth corps?

I think they’re saying that these small teams getting screwed aren’t exactly going to win a legal battle against Meta and Google, even if they could afford to try.

Google is notorious for over-reporting ad data for all sorts of reasons.

One super important thing to know is conversions could also be landing page views depending on how the ad account is setup. By default, Google will use any type of conversion configured in the ad account for displaying stats in the campaign. It's a bit tricky with Google Ads to configure a specific campaign to only report conversions as app installs, signups (leads), or sales.

Best practice is to mostly ignore the data reported in Google Ads dashboard (unless you really know what you're doing) and instead rely on your own metrics. Run multiple campaigns to different audiences and different landing pages to be able to more easily verify what's actually working. Fortunately, it's fairly easy -- although not straight forward -- to pipe Google Ads data into a Google Sheet and merge it with actual data from your website. I'm not sure how this works with pre-registration ads since I've never run that specific type of campaign.

This seems like it should be a lawsuit waiting to happen. If I was running ad spend through Google and they were giving me inflated reports routinely, its willful on part of the company, and feels like grounds for a lawsuit.

The problem is that it's one of those things that to take it to court, you need some starting proof, but the entire system is designed to be opaque.

It's not necessarily inaccurate data from a legal perspective. But it's often unexpected data where it's getting reported in a less useful way in the dashboard.

And yes, there are almost always ongoing class action lawsuits agains major ad platforms for failing to properly filter out bots and click fraud. There's very little incentive for Google and Meta to do more than the bare minimum in fighting bots.

I have built ad network tech from scratch and worked for several large networks.

Here is my tip. Always hire specialists to do these sort of ad campaigns for you. It is not something you can figure out easily and lessons here are expensive. OP learned some of the very basic principles of this industry for a price tag for $20K. I have seen people do even worse.

1. Whatever Ads are one of the options you have. Other options are paying users cash to install your app, billboards, tv ads, radio ads, asking teens to distribute app install stickers and what not. I would recommend trying the cheaper method first that do not change much with scale.

2. Ignore the network reported stats. What matters is your business. What is the true cost of acquiring a user that meets all you requirements ?(e.g. paying user) In OPs case this number could be in $00s now. Totally not worth it.

3. Learn that number with smaller spend. (Smaller spend = around $2K for Google)

4. Optimize on that number using different strategies.

Chances are that for most apps your cost to acquire a user would always be more than the revenue generated by that user. Making these ads always a losing proposition.

Then why do folks do it ?

* If you spend $0000000 at scale (like Uber, FB in early days) even though you are making a loss per user, the large number of users with network effects snowballs bringing organic growth. This is compounded if the app itself deals with real money (Uber ) or has deep network effects (Signal, Instagram, Snapchat) which bring in organic growht.

* A meal prep app is never going to be profitable in this manner.

If I was OP I would have probably tried to find all the instagram influencers who talk about meals and then emailed them about a partnership. Would have paid $500 or so to tier 2 influencers to make coiple of posts about the app.

On other than those $20K helped you get this post on the front page of HN. This is going ot add few hundred installs for your app. Note that you salvaged that damage already.

16,171 converted and pre-registered for the app

1,371 either installed it when they got the push notification or selected auto install when they pre-registered (it's oped in)

Cost was $22k USD ish.

This is extremely normal... it's almost exactly as I would expect, maybe even good! I'm curious why they would expect differently?

If you didn't tag your app as having Ads, or having in-app-purchases but it is tagged later or released that way all the pre-registrations that happened while not having the appropriate will not be able to auto-install.

For legal reasons, the user needs to know what it is signing up to at the moment that agrees to install (auto-install = deferred install). I think that makes sense but obviously apps become less attractive when those tags are displayed.

I also think the pre-registration program makes sense for games rather than apps.

I agree with the author's assessment that only 1371 users were actually converted and they were charged inappropriately.

This complaint is tantamount to me saying: I spent $20,000 advertising my new cloud provider on twitter, 5000 people signed up for an account but only 5 of them ever logged in so only 5 converted. That isn't how it works.

In google ads a conversion is clearly defined as the users clicks the link and then completed a second action (pre-register in this instance). When you create a campaign, you define the goal like this: when a user clicks the link and pushes the pre-registration button the user goal of conversion is complete.

Just because you don't like what happened thereafter isn't the concern of the ad platform.

To take this even further, the install isn't what matters either, it's the user buying an in-app purchase, or it's the user still being subscribed 1 year later, or it's the user buying high margin merchandise after 3 years as a loyal customer...

Every part is a step in the funnel, and every step down the funnel is harder and slower to analyse. Ad marketplaces only have the data of the very first step most of the time.

Doing this analysis is The Job. That's all of what marketing (particularly digital marketing/PPC) is, and being good at this and building the loop is the difference between buying clicks from people who aren't actually that interested, and buying clicks from people who will pay for a product.

It's because Google clearly wants it to seem like pre-registration is actually pre-installing when in reality is closer to pre-nothing.

No, it amounts to Google claiming that 13k users will install the app and billing for their hand in that when the actual install base was 1.3k.

With 17,000+ pre-registrations, it sounds like their ad campaign wasn't well targeted either. You can get a ton of people visiting your app page but if it's done based on vague or generic keywords, the clicks you get won't mean much.

Seems a bit crazy to spend that much without any cap for a small indie company too. If the app is any good, you can always run more ad campaigns after the launch.

We didn't cap it because we were getting conversions at around $1.36 USD. This is more than we'd normally spend for an install, but we thought it's worth it because it also gives us launch day momentum. Obviously, had we known that actually only 1 in 12 were going to install, we'd of killed that campaign instantly.

Thanks for sharing your experience. I would have thought they would auto-install the app for users that pre-registered, as Apple does. Seems kind of pointless that they don't do this!

Hope it's helpful for some fellow developers when considering how to promote their app launch :-)

For a sense of scale, if an advertiser came to my adtech platform and said they want to spend $20k, we’d tell them to return when their budget is 5x larger.

Small advertisers get screwed because they lack the technical expertise to understand the complexity of adtech platforms. Those complexities provide huge power and value to people who know what they’re doing, but also cause huge problems for advertisers.

I don’t know what happened in this case unfortunately.

I think they're being a bit misleading - the issue is simply that Google counts a "conversion" as someone signing up for pre-registration, whereas they assumed it would be for an app install.

Google should be clearer about that, but it's not like they are just making numbers up.

Fair point. To Google conversion = someone who has pre-registered. That's different to someone actually installing. I never expected 100% of people who pre-registered would install, but having only 8% of people who pre-register go on to download seems kinda wild.

If anyone has any figures for a pre-order campaign on the App Store it would be very interesting to compare the conversion rate.

Hopefully someone with some benchmarks on this specific kind of thing can chime in, but to me it seemed like 8-10% conversion from notification to install was not too terrible. Don’t get me wrong, it sucks that it turned out this way! I’m just basing this on conversion percentages I’ve seen from email launch sequences (maybe 2-5% of the list), webpage conversions (free signup maybe 10,20,50% but paid 1-5%), and 8-10% for a nearly-cold install notification seems kinda in the ballpark.

There’s a correlation I’ve noticed between how easy it is to sign up for a thing, combined with how badly someone wants that thing, and then how likely they are to actually buy/download/etc. In this case it was super easy to click that “pre-subscribe” button so I’d think those people aren’t very invested. There’s only a tiny bit of time for them to form a connection with it, so when that notification pops up hours/days later they might not even remember they clicked that button, or don’t remember the name, etc.

It’s a bummer it was so expensive though! I went through a couple expensive YouTube paid promo experiences recently that took me by surprise too. Just like nowhere near the performance I hoped for. Not fun!

What was painful about pre-registration ads is you don't realise your mistake until you launch your app. Only then do you look at your screen and then go have a cry hahaha.

Be super interesting to read a case study for the App Store and their pre-order ads.

I think the problem is simply that the author misinterpreted the term "conversion". A conversion is not an app install, is a successful pre-registration.

Now, it seems like there is a 10% conversion rate from the pre-registration to the installation of the app. Why that is there? That's a different question, and probably explained by their docs regarding pre-registration and what it actually means.

$10-$15 for an app install is probably pretty normal.

The app seems to be running at 60hz on Android with 90hz screen. It's a common giveaway for Flutter apps, but can be fixed by forcing higher refresh rate.

I spent some money on GP ads and it was kind of funny that no matter how the views and clicks were distributed in the end the cost for each install was about 1$.

It is "fake" how it works but it is in the agreement. Also Google ads has the worst support ever. Not even helping when you want to invest.

Fun fact. GP actually paid out some cash invested for ads this year. Have no idea why. I am pretty sure I burned it all on ads.

Is this google ad campaign linked to the play store directly? How does the ad campaign figure out whether a 'download' or conversion happened? If its a click/google search campaign, won't it stop at redirecting a user to the app download page?

If I am understanding this right, the crux of the issue is that Google Play doesn't auto-install apps that you seemingly signed up to have auto-installed?

Seems like a bug on Googles side, whether intentional or not.

If it's a bug then I'm hoping to raise some awareness of it here because the Google support team didn't want to hear about it.

Just out of interest, did you do some social media stuff? You can get quite some traffic by presenting your app in a relevant subreddit. Reddit ads are also much cheaper.

If you are like me and confused about how pre-registering doesn't cost anything but downloading could be triggered automatically on the launch day: the app is free.

Free is the default price for mobile apps. You're dead in the water these days if you have any sort of paywall blocking the user and a download. they will just find 20 other alternatives without that barrier.

We will do very well as a purely digital world: everyone thinks they make it because they have an edge while realizing to late it's not.

What a naive take on using 20k

Thanks for sharing this. I was under the impression that Play Store pre-registration results in automatic installation. This is good to know that it is not.

Sorry to say but there is no money in this industry/category. I know from experience. Get out, build an app in another category.

Thanks for the heads up - I won't use this for my project.

Happy I saved you some $$$

Wow this is crazy!

Yeah when I first looked at the analytics in the Google Play Console I thought there must of been some error. Figured it would be a fairly quick thing to sort out with Google support.

Sooooo.... Fraud?

Hm, what's the point of placing TLDR section at the end of the article? :)

Hehehe good point. I removed it and linked to the HackerNews thread instead as that might be more useful for people.

Reply on news.ycombinator.com