How we unnecessarily burned 20k with Facebook Ads | Hacker News Reader

How we unnecessarily burned 20k with Facebook Ads

stackfield.com

280 pointsrolfos116 comments

> Even these users churned after a short time. Why is that? We do not really know!

You need to know this, and knowing it is as easy as asking.

I e-mail every single person that cancels their subscription, and ask why they cancelled today and what we could have done to better serve them. I don't measure the response rate, but I write every mail personally and read every response, so I can tell you it's well over 50%.

The responses to these e-mails are invaluable. They're how you learn why people signed up, what pain points they experienced using your product, what problems they have that your product didn't solve well, and what competitors they're coming from or leaving to.

If there's some glaring issue with your ad campaign ("I thought I was signing up for a free antivirus program because your ad said it made my files secure" or "No hablo ingles") you'd probably have learned it well before spending €20K through those exit emails.

I am curious how you have used the information to change your product to reduce the churn? What are the results? How would you scale this if you had a larger customer based?

(Separately I am always amazed at how larger companies do a lame attempt at the same thing by expecting me to fill out a long survey with simplistic categories after a luxury purchase where they could clearly afford to give the person touch and show actual interest in my answers).

I can't speak for them, but I've worked with exit survey data that has either confirmed what we knew about some market segments we weren't going after, or highlighted some unforeseen pain points in the product that we later scheduled to fix.

And you're right - the longer the customer survey, the less genuine the company is in actually hearing what you have to say.

Our experience when trying this approach is 90% reply with 'lack of time' which is not the true reason but basically means the product didn't solve a major problem for them.

Not necessarily. I solved most "lack of time" issues by improving onboarding. People thought they didn't have the time because the setup steps seemed daunting or confusing, so they put them off indefinitely. Improving the "first login" experience, moving to a more walkthrough/wizard-based setup (only one thing to do on the screen at a time), and building more 3rd-party integrations that eliminated technical work allowed me to retain more customers. I was solving their problem -- once they got past the hurdle of starting to use the product.

Absolutely. I'm a picky user who will delete your app in 30 seconds if the setup path isn't obvious and reliable.

I think you're probably just a reasonable user.

Yes. If a company doesn't even manage to optimise the onboarding experience, it's unlikely that they optimise anything.

It depends on your audience as well (surprise, you need to understand who will be using your product). What we've found through AB testing with the game I work on is that our best (read: highest-value) players are very experienced with this type of app even if they're not tech savvy, and a tutorial is a nearly meaningless roadblock to them because they already know.

To clarify what I mean by "doesn't solve major problem": it actually might solve a major problem for them but due to shit messaging/ UX/ onboarding/ trust/ bugs/ whatever they haven't realised it.

So you are correct but all those things I just consider to be part of the product as a whole.

> which is not the true reason

Wait, how do you know this isn't true? Maybe your product had so much friction that they just didn't have the time to learn it, get on boarded, etc. Seems weird to dismiss approx 90% of feedback you're getting from users as being lies.

It's not a lie, though.

It just means they saw some value in the product but didn't feel enough motivation to actually follow through and signup/use the product.

This is similar in many ways to someone answering "it's too expensive". The value prop is either not strong enough, not being marketed well enough, too complicated to start using... or the product is solving a pain point that people don't care enough about to build a business around it. Which is a serious question the OP needs to ask themselves. Unless their traffic<costs>conversion rate ratio is sufficient.

You can't expect customers to articulate that everytime (10-20% seems about right to me) because most people aren't actually thinking their actions through when visiting a landing page. Especially when it's via an ad... not something they searched out on Google or via referral.

Even if they did tell you they wouldn't necessarily know why... because you're assuming it was a rational thought process, or something they spent time thinking through, which is rarely the case. Books about how customers acting 'irrational' [1] are very useful at explaining this and are important to always be considering when doing marketing/UX design.

[1] https://www.amazon.com/Predictably-Irrational-Revised-Expand...

We definitely do NOT dismiss it!!! It just means you need to spend a lot more time finding the real reason, or maybe you are getting traffic from the wrong group of people who have no interest in your product anyway.

Good example would be Product Hunt or Techcrunch as a traffic source, mostly useless for any testing as everyone on those sites is just interested in 'new products' or startups and is unlikely to be in your target demo.

> true reason but basically means the product didn't solve a major problem for them.

How is that not a true reason?

If a product isn't valuable enough for users to spend time on it (or it doesn't save them time), you don't have product market fit.

>, Facebook may just be the wrong audience for a B2B tool.

Well, at least you can say that the 20k € was spent to learn that signups from Facebook have a near-zero retention rate. (That's not meant to be a flippant comment.)

A lot of successful websites seem to know (or learned) which market channels bring in the right customers.

For Stackoverflow.com: most traffic comes via referrals from google.com searches -- but not facebook

For Craftsy.com: it's referrals from Facebook -- but not google search queries

For a project management tool like yours, neither Google Adwords(1) nor Facebook newsfeed ads will work. Maybe a strategy such as hosting for free a high-profile open-source project so your brand gains mindshare. Then it may attract non-open-source paying customers.

That subsidy strategy for non-profit customers is not foolproof (as it didn't seem to help BitKeeper) but it did seem to work for Github.

(1) I suppose you could try running an AdWords campaign on keywords such "project management" and related competitor keywords such as "Jira" and "Microsoft Project" -- but I doubt you'd get quality clicks from it

"but I doubt you'd get quality clicks from it"

I don't think it's the quality that will be the problem but the cost in a highly competitive area like project management.

Their problem is they are B2B, not B2C, for which social channels tend to be better at. Regardless of traffic, they should always think about running search ads for their branded keywords at least.

Also their differentiation from the rest of the market is E2E encryption, for which there is a small target demographic within project management. Like you said, rather than rely on advertising channels they need to offer functionality to stand out from the crowd, which should be useful but not give away their core revenue services.

They're already starting to do the right thing - Blog posts demonstrating they are subject matter experts, that can get exposure on places like HN (which tends to have demographics who care about E2E encryption).

A lot of successful websites seem to know (or learned) which market channels bring in the right customers.

Yeah. We do grant writing for nonprofit and public agencies (see http://www.seliger.com if you're curious), along with some small tech companies, and for us ads that target conventional consumers are useless. We did find Google Adwords useful for many years, but over time its effectiveness has declined to the point where we stopped using it. We've tried some advertising in other domains but found them ineffective too. A lot of our traffic actually comes from search queries that pull up blog posts.

I think the larger point is that what works today may not work tomorrow.

I find it amusing that a product focused on privacy choose to advertise on Facebook. I bet that being on the front page of HN is well worth those 20k though, but maybe that was the plan from the beginning.

Are the assumptions here that:

1) Facebook News Feed users don't care about their privacy

2) The OP was more motivated by viral HN marketing than actual direct results on Facebook?

These are interesting claims. Can you substantiate them?

There is an enormous difference between a user that is searching for keywords related to your product, and potential customers that probably aren't looking for your product that you have identified through Facebook targeting criteria. The Google user has a problem and is actively seeking a solution; the Facebook user fits what you believe is the demographic profile of someone that might be interested in your product.

Traffic from Facebook ads is notoriously difficult to turn into cash precisely because of this. Most of their advertisers measure ROI in terms of shares and likes, not new users or product sales. Because Facebook has you bidding against such people for traffic, you will never be able to make the numbers back out because your competitors aren't measuring dollars or new users like you are. Advertising products/services on Facebook is a fool's errand for almost everyone, except for those with an extremely broad target audience.

That's not the case at all. In fact, advertising on facebook (at scale) requires very complex measurement with statistical models.

People need to understand that marketing (in this case advertising) is hard. Just because you can code or do something else and you are a generally smart doesn't mean you can figure out how to buy advertising after 20 hours and reading some blog posts straight from google's search results.

Dear god this. For most advertisers with a decent budget, measuring ROAS for FB on a last click basis across channels will lead to disappointment.

Unfortunately, developing a proper attribution model with post impression conversions (not just click based) is arguably the hardest challenge in the industry today. There is no one size fits all approach and even as someone who does this for a living and is well versed in the matter, I still haven't found a great approach. Incremental list testing isn't always easily done.

That's the essence of search vs. display advertising. You're right that the user's mindset is totally different when searching vs browsing but I think your conclusion about display ads not being worth it for direct online sales is wrong. People make money selling goods and services through FB ads. And many people are profitable with this (in the classical sense of "ROI"). Perhaps you're in a market where you found yourself bidding against people who don't measure direct monetary returns, and yes this is much more prevelant in display as opposed to search, but it's not that polarized in general. Your click through rate and conversion rate will be lower. Your cost-per-click will also be lower. Display works for direct advertising in many industries as long as you know how to target your market's interests.

Also fwiw I haven't seen the term "ROI" being used to describe non-monetary returns (clicks, likes, etc...), is that common now?

Re: ROI, I was just generally referring to the fact that many advertisers, large and small, deploy campaigns with the goal of gaining likes/shares etc. Brand advertisers count ROI this way, along with small businesses that have no other way to measure ROI other than with metrics such as likes, shares, and follows.

BTW, I am not the only person that feels this way about Facebook ads. I can't find the link atm but a study I read a year or so ago indicated that less than 6℅ of FB advertisers are showing a positive (cash) ROI.

"The Google user has a problem and is actively seeking a solution; the Facebook user fits what you believe is the demographic profile of someone that might be interested in your product."

That's all well and good, but at the end of the day I can Google "Maserati Dealer" and yet have nowhere near the income levels required to purchase one.

"Traffic from Facebook ads is notoriously difficult to turn into cash precisely because of this. Most of their advertisers measure ROI in terms of shares and likes, not new users or product sales."

As someone that works in the industry this patently false. We measure Return on Investment in dollars not vanity metrics. ROAS is going to be your most common topline stat which is used (stand for return on ad spend with return meaning revenue).

"Advertising products/services on Facebook is a fool's errand for almost everyone, except for those with an extremely broad target audience."

Again very very wrong. In fact the most successful campaign I've ever run (3 million+ in revenue that was 300% ROI positive) was a relatively small niche market that was less than 200,000 people.

Now if you want the real reason Facebook seems to not work. It's a few things.

1. The people that raise their hand up to talk about Facebook ads do so because they're frustrated so all the things you read tend to be very polarized and negative.

2. Mostly amateurs are buying Facebook ads. Professionals routinely do much much better even in B2B niches.

3. Just because you're product didn't sell doesn't mean that the platform doesn't work. It routinely works for thousands of advertisers of all sizes every single day. I would encourage you to ask if it's the ad platform? Perhaps your ads? Perhaps your targeting? Perhaps your product just hasn't found product market fit yet.

You seem to be assuming that my comment was the result of some failed campaign. It wasn't.

Surely you would concede that your $3M campaign was a rarity and could not be duplicated in the case of most advertisers. About 30℅ of casino players walk out winners on any given day too - it doesn't mean they are going to beat the house long term. If you are advertising on a medium where useless metrics are the desired goal of your competitors (likes and shares in the case if Facebook), you cannot beat those people if you need a cash ROI and they don't. There are always going to be exceptions to the rule, and I'm glad you found one, but that doesn't invalidate anything I've said.

AdWords has demographic targeting and in-market segments, you can be slightly more targeted than broke people looking for Maserati's.

Agree with your general sentiment though, it's not necessarily a platform problem, the real problem may be expertise or a meh offer.

I wanted to call this out:

We solved the first problem through a separate tool, which labels the user with the originating source when registering.

Yes, yes, yes, and yes. If you're spending any significant amount of money on advertising, you need to be doing this. Don't rely solely on third-party cookies and session unification attempts. If you have a USER table in your database/crm, add the five UTM fields to it! Conversions for most businesses are a multi-day and, often times, multi-device process.

Can also just add a field to signup "how did you hear about us?"

Or both, I was dubious about a request to add this field where we were doing proper ad source tracking via URL variables, but it led to some interesting results, specifically people coming via Facebook ads and the lack stating word of mouth, or offline advertising, as the source.

If you're using a tracking tool like Mixpanel, you can get the UTM parameters for each user and enter them as hidden field values into whatever form they submit so that you have them in your DB/CRM/email.

In truth, you should be doing this and modeling the results via attribution for every source of traffic. What if the facebook audience bounced right away and converted via google two days later?

Disclaimer: I'm working on an attribution modeling tool, attribution.io

I was really interested in your product.

Unfortunately your website was less than informative. Any chance you could provide some more details around your approach and methodology?

You're right, the landing pages are basically just teasers to get people interested. It's very much in beta.

Happy to provide more details. The platform works by tracking the visits to your site and determining where they came from (via referrer + url params / utm tags). Then when an event is fired, I look back at the history of that user's visits and run attribution models to distribute credit for the event back to the sources.

Technically, I'm using a few rails sites, sidekiq workers, redis and postgres to keep it all humming. Not much fancy on the front-end, just plain ol javascript.

Is there something specific you're interested in?

Edit: feel free to mail me if you want more personalize info / discussion: hello @ attribution.io (added spaces for spam robots)

Are you doing anything with dynamic/data driven models? Or do you just focus on the same static ones as GA plus your static content models?

Also, how do you handle view through and other harder to measure branding touch points? Do you integrate with things like DCM?

My beef with most non-enterprise solutions (which I'd classify yours as based on pricing) is they are very limited, and typically don't do much of anything beyond what is available for free in GA.

Fwiw I'm a senior digital media guy who goes deep on attribution. It is what keeps me up at night as we work on display,video and paid social. The fact that we have a longer sales cycle due to a free trial further complicates things.

I do have a few questions...

Is there a way to do TV attribution beyond multivariate statistical learning models attempting to gauge the impact on other channels based on broadcast period? (i.e. treat the offline channel as a categorical variable)

What is "dynamic" attribution modelling? Does this mean the attribution model changes depending on who you perceive the customer to be?

What is the best way to do multi-device tracking if you don't have a login?

I'm not who you asked, but maybe I can help with my viewpoint.

TV/Radio/Offline marketing can only be attributed via these statistical models (eg in the zip codes where the ads ran, did a bump in traffic or store visits happen). Although old fashioned, I think coupons per channel are a great way to get confirmation of the link between ad and purchase, it's less precise than digital tracking, but at least makes some concrete connections.

Multi-device tracking without a login can only be done by using a huge network of 3rd party data. E.g. if xyz company has cookies on everyone and they see a login on one site, then they equate that cookie as the same user across devices, effectively sharing the information with their clients. I think this is very shady and I do not use any data like this because I think it violates people expectation of privacy across interactions.

Have you tried correlating visits from different devices from the same IP? E.g. if you get 2 devices from 3 St John St within an hour, then the same 2 from The IBM Building again within an hour, you can make a fair assumption that's the same person (you need the second site to isolate family members).

I've never implemented this - never had the time nor clients who wanted it enough - but have always wondered whether it might work. Maybe later this month. In declarative SQL, naturally.

Check out what Adometry and other dedicated attribution providers are doing if you are interested in TV. You need a hefty budget and volume to use them though.

Basically it is a combination of art and science with the statistical modeling being more on the art side, and then the science being tagging the ever-loving crap out of everything you can through a single platform to record as many touchpoints as possible.

TV is increasingly becoming measurable through things like Hulu, etc. You can also do vanity URLs for direct response TV ads.

Dynamic attribution modeling uses machine learning to continually evaluate the effectiveness of touch points, and then changes those weights over time. There are soooo many variables to factor in (decay rate, creative, placement, etc.) and every user has their own unique conversion path where each touch point might have a different level of impact. It is in theory the holy grail of marketing, but totally unique to each company and VERY hard to implement decently.

Multi-device tracking has come a long way thanks to lots of data partnerships, but honestly a lot of that feels icky to me. They are effective though. A lot of them rely on cookies and using data brokers and cookie onboarding services to have cookies continually re-added and matched across various devices. So you may not login on my site, but if you logged-in to another site that partners with a data broker (or is Google/FB who do this on their own), they know who you are, and can then map that data to existing audience data records.

All of my models are 'data-driven' in that they are computed directly from the visit & event data. I don't currently have models that change based on the data, but I could easily add them if you're interested in helping me define how they work.

I find one of the big problems with GA is that it doesn't show you the actual people that are included in the final attributed credit per channel/campaign, does it? Without this, it's quite a leap of faith.

I'd love to chat with you about your current practices; I still think there's still tons of room for improvement past the current models. For example, I'm experimenting with a model that gives credit to pages in addition to traffic source to help justify the cost of content production / optimization of which content is working and which is not.

I'm using "data-driven" synonymously with "dynamic" in this case, which is indeed the changing model based on the data using machine learning.

I haven't played with it to this extent, but can't you get a GUID of some sort using the GA API[1] or custom dimensions? I haven't tried it personally and I know they restrict some of the data so this may not be possible.

I do like that you are factoring in content into this. The regular way of doing that is to create some custom channel groupings in GA (probably some other ways) but content really deserves a place in the funnel when looking at attribution. And you are right that content marketing is hardly free and has a high cost associated with it that is hard to prove ROI on.

Definitely would be open to chatting. I work in Mountain View if you're in the area and that would be easier.

[1] https://developers.google.com/analytics/devguides/reporting/...

You can also ignore UTM terms altogether, and use a single ad ID, then structure your marketing channel/partner/etc. exactly how you like it at DB level.

Aside from being more correct, as you don't duplicate this information with every session, you can update your "UTM parameters" at any time (keeping a history table, naturally) and avoid enormous and growing sets of rules to map previous structures to the newest one.

Totally agree. I literally just spent 9 months getting this integrated at work - now everyone sees it as magic. For people with tech and marketing in their arsenal it should be one of the first steps if you're working with standard tracking link variables. Obviously (with our set up) we're dealing with last touch attribution, but it's still infinitely more worthwhile than not having it.

How can this can be achieved this for mobile apps?

Localytics and Mixpanel can do tracking across platforms. And I'm sure there's more.

Not only that, but also add current_* and original_* variations for those fields. Sometimes you have channels that are good at bringing awareness but others that are better at converting. They are not always the same.

This is basic advertising. Google search ads and referral will bring in higher quality users because search is trigger/intent-based and referrals come from a more trustworthy source (usually), vs. FB ads that basically blast everyone. Even if you narrow down your audience on FB with interest targeting and proper demographic targeting, it may not help since many users just 'like' everything. Consider the supply and demand as well, the less competition on FB the cheaper your advertising is. Perhaps competitor also realized FB was the wrong way to go and stopped advertising on FB, that means more auctions for you to win and cheaper users. Of course, that doesn't mean higher quality users.

In my experience, facebook is for marketing ( getting your name out).

Google is for results

( We had a facebook page with Pokémon Go, where we had 6 events in theme parks. We did it all with facebook and a website... None the less, i seriously changed my opinion about Facebook considering ads. It's not worth it )

We had a reach of 320.000 people with only 100$ spend on ads ( because of the hype it was really cheap and a lot of people shared everything).

We could only convert something like 300 people per event.. Which is 0,00009% conversion rate, where they pay lower entrance fees then normal for a theme park, in the midst of a hype.

Remind you, the first event was in the first week after the release of Pokémon Go.. So we really jumped on it :)

PS. We spent more on ads for

- like our page

- better position of our likes

- ...

It's just not worth it

300/320.000 is 0,09%, and not 0,00009% .

Yeah sorry, posted it too fast obviously

I recommend you don't use the conversions/reach number as your primary KPI. From my understanding, you spent $100 on ads, and got 300 conversions.

If a conversion isn't worth $0.33 to you, I would recommend re-examining your pricing structure and perhaps even model.

I may have misunderstood what you said, in which case I apologize. My point is not to criticize your effort, or deny your experience, but to correct a perspective I see a lot.

Don't worry about the number of impressions. Focus on how many customers you can get, then back out the numbers to find out how much revenue that made you.

I am not saying it wasn't a succes.

But the success was because of the hype, not because of Facebook. We could have pulled this off on other platforms as well, we were 3 times in the news.

But the success isn't as big as it should have been. We paid pennies and had something. But I wouldn't want to know how much we should pay when I'm an actual business and trying to get some traction through Facebook

We could have pulled this off on other platforms as well

I have to ask: why didn't you?

Because this seemed the easiest way and i was curious about facebook for my ( now already launched) webshop.

Now i just consider facebook as marketing/brand building and not a big part of my sales funnel.

Seems to be correct in my experience so far, most of my sales currently come from offline and 3rd parties ( one friend of mine brought in a huge sale)

+ I only have a limited time :)

One thing I have learned in SEM and paid social -- it's difficult to extrapolate experience across projects.

I've had projects, appeared nearly identical, similar ad copy, targeting, etc. One was wildly successful, one was a flop. Going in, I remember assuming the second would be a big success, but it absolutely wasn't.

There is no should, there is only what happened. It's very difficult to project with sufficient accuracy, the variance swamps all.

"Facebook might be the wrong audience for a B2B tool" sounds most likely.

I think so too. FB seems more appropriate for B2C products, particular ones that are commodity.

I had a bad experience with Facebook Ads. There were a lot of people that were liking the advertised page, but at the end of a day, all of these people' accounts were inactive. They never actually liked or shared anything published on FB page they liked before via Ads. I have a feeling that Facebook is generating fake likes, trying to make an illusion that the campaign goes well, i.e. pushes people to spend more.

Facebook ads are primarily great for three things; selling games, physical consumer products and local services

So if you are selling mobile games, t-shirts or offering plumber services or are a dentist it's great. For most other things it's not really useful.

We tried to use it for local services and we failed. All of the likes were coming from people who did not have anything to do with our service. It was really bizarre. I made the target audience very narrow but could not get a single hit that was a real potential customer. We switched up strategy and just asked everybody in our social network to like our page and share our service. This helped quite a bit, even though this is for free. :) Next thing we did was using meetup.com and target the right meetups with a discount campaign that worked out pretty well, this was also for free.

I have had success for those I have helped. My dad picked up quite a few jobs (he is a plumber by education) that way and we only spent around 100USD.

But I am sure there are things where it doesn't work or areas were it wont work.

With regards to asking network yes, thats the first thing I always ask them to do.

In fact I don't understand why larger corporations do not have some social media incentive program for their employers so they can help spread the word and perhaps get leads.

I put a lot of money into advertising through the "proper channels" for my business, including Facebook Ads. We were crushed by a competitor who activated a few links within his PBN and spent some serious time astroturfing, both tactics that I had refused to engage in.

The sad reality of online marketing is that people who are buying ads are suckers. The only way to be competitive online is to actively game the mechanisms that the few big traffic brokers use to determine rank.

What is supposed to be astroturfing in this case? Infomercials in media?

I'm not sure I understand your question. Astroturfing is getting online and posting positive stuff about the product while posing as many unaffiliated persons on forums, review sites, social media, etc. This creates an appearance of high popularity and user satisfaction, and like astroturf, you can't tell it's "fake" from a glance. This creates a buzz that causes real users to pay attention and join in.

Since this company was targeting Facebook, if they wanted to astroturf, they'd do it by taking control of many Facebook accounts and posting positive stuff about their product in niche groups, on the product page, etc.

You understood the question perfectly. Sorry if I wasnt't clear enough. Thanks for explaining it to me. Call me naive but I didn't know they go to that length. I am aware of usual infomercial, PBN and outreach stuff but building the whole armada of of fake identities.

I am all for playing clean but in this case it seems there is no other way than joining this train(wreck).

So many people using online advertising tools face similar issues. As many said here, setting the right success metrics + using the right tools to measure them is key for a Facebook Ads campaign to work well.

The first thing to understand is how different Facebook metrics are from your metrics.

Clicks on Facebook Ads are not visits, for instance, and the drop-off you're seeing is just normal.

Then, Facebook allows to track many events that can help you acquire the right people, including the one who stick, whether your product is B2B or C. There are 1.7bn people on Facebook, meaning getting the right ones is more than possible.

Experience is key, I work in the field and I can already tell you that your ad and bidding options were probably not good enough to get the right users. You basically asked Facebook to drive traffic to your website, never to generate long lasting customers... so to me, Facebook did what you asked for.

If you want to be successful on Facebook, you must use their Pixel, and optimize towards events that define success for you.

Place the Pixel on your site now, even if you don't plan to buy ads on Facebook right away. Set some standard events on key pages (e.g. purchasing a subscription or similar). This will help you for the next time, as you will be able to create lookalike audiences based on your best users. This audience, ANDed with some interests will allow you to target your ads to the right people. Combining this with a good ad, and a good bidding option (e.g. oCPM optimized towards subscriptions) will probably lead you to posting much better results on your blog next time.

Lookalike Audiences: https://www.facebook.com/business/help/164749007013531

oCPM: https://www.facebook.com/business/help/494633817315490

>Surprisingly, Google Analytics reported only 28.818 visits by users who were redirected from Facebook during this timeframe – i.e. 8,81 percent less. From these users, 3.786 registered on Stackfield. //

Hypothesis: Could that be pre-emptive caching - some browsers IIRC will fetch a page ready for display. GA could be showing displayed pages whilst FB are showing followed links?

On the general situation, what's the rate of false clicks on Facebook - like how many people accidentally click an ad?

Do browsers really preload pages from random links in advertisements that I have no intention of ever clicking? That sounds like a terrible idea.

I don't know about that, there may be extensions or add ons that provide such features though.

Many browsers will start loading a URL as soon as you type it in, even if you don't press enter however, try this with a local server sometime, it can be a little jarring.

Step 5 here - https://support.google.com/chrome/answer/1385029?hl=en-GB.

Or https://developer.microsoft.com/en-us/microsoft-edge/platfor...

Looks like Edge only prefetches when specific link tags give prefetch prompts (Resource Actions per W3C specs) whilst Chrome appears to do things more speculatively.

Not sure about IE, FF and other approaches.

Google Analytics doesn't report exact results (unless you pay for it, around 150k I think), 10% of error could be normal. It should be use to measure trends and not exact numbers.

8.8% fewer clicks in GA than what you paid on FB isn't bad, IME.

I'm running a campaign where FB charged me for > 800 clicks. My GA system recorded about 300 clicks.

Slightly under half is about expected whenever i look at numbers. I'm doing remarketing to people who visited my site in the past 7 days, so I'm not sure if that has something to do with it.

My biggest problem with FB advertising is that it is virtually a casino. We do everything right (I think). A/B test, segment, measure, measure, measure. But somehow it's never reproducible. A great organic post can flop completely when advertising to the very same target audience.

We're B2C selling a targetted learning subscription to students in the medical field early in their learning path. We don't have fierce competition at all. Facebook should be the right tool for us. Facebook did work great for a while, but suddenly doesn't. Or maybe it would tomorrow. Who knows.

We're still spending money and trying to optimize, but somehow it always feels like we're basically gambling the money away. We some times get lucky, some times not. The house always wins.

Try Instagram and SnapChat. Facebook is still great but it's hugely saturated. Instagram and SnapChat seem to be where all the brands are heading now. Also Musical.ly but that's if your target demo is teenagers or younger.

If you have the budget and persistence, you need need to be doing retargeting. Sponsored posts are... meh.

We are doing retargeting as well. Not just sponsored posts.

I think one of the main things people forget in B2B stuff is that impulse purchases are less of a thing and people want to spend more time evaluating value/spending. This is why many a lot of B2B folks setup nurturing campaigns.

This is where Facebook can really shine - what if instead of pushing a hard sell when someone is in the middle of reading their friends updates, you promote a white paper to folks that are already in your campaign? Or give them something else of value, but make it about them, not you, and not about hard sales – that comes at the end of your nurturing campaign, which should not be on Facebook.

This! I forget the company name but one was on This Week in Start-ups a while back and they have a funnel where they use their data on you and, if they found you started looking at their products, their ads would change to something like "come on back" and if you started building a couch it would change again to "you can have that couch delivered in a week!" or something similar to that.

It's a little creepy mind you but I think the vast majority wouldn't notice that and it's a huge way of getting people back in.

This is a valuable lesson in tracking marketing results.

Decide on what would make a campaign successful, and then make sure you can track it.

Don't trust the ad platform's metrics. Not only are they inaccurate, they're usually vanity metrics that don't matter at all.

Tracking outcomes can be as simple as including UTM tags in your ad links (eg, ?utm_source=facebook&utm_medium=cpc&utm_campaign=fall2016), tracking signups/activations/upgrades in Mixpanel (or Heap), then viewing a report of signups/activations/upgraes grouped by the UTM values.

>> Why is that? We do not really know!

So ask... You have a few thousand users who bailed. Drop a survey on a reasonable-sized random segment and find out why they left

So your churn of users acquired through Facebook ads was almost 100%? To confirm my understanding, thats why it was wasted? Why do you think they almost all left? I don't see a free pricing tier on your website now. Do you think the free pricing contributed heavily to your churn?

I see a "Try it free" link on the bottom of the page.

I'm just about to start my first round of facebook ads for my b2b startup. I was wondering what worked best for you? I read your post and you mentioned users from sources such as Google... what else worked for you?

Don't bother with FB. It's a waste of money for B2B.

Nothing is guaranteed to work. AdWords could be a great way to jumpstart customer acquisition, or a great way to waste money.

The important thing is to _experiment_ and _evaluate_ (using analytics and attribution tracking) because what worked for one company might not work for you, and as a startup you often can't afford to blow $20k just to learn that lesson.

(Speaking from experience. I'm a b2b marketing consultant and I've been through this many times.)

Would you recommend hiring a b2b marketing consultant ;)

Just like with everything else, if you have the time and skills to do it yourself, then no. Otherwise yes, get help. If you know exactly what needs to be done and how, then hire a freelance specialist. If you know something needs to be done but not sure what or how, then hire a consultant.

If the poster is here, may I ask – who does your illustrations?

A quick Google have me this link : https://dribbble.com/stackfield

It's the CEO. Wow.

Missing one of the most useful metrics and that is CTR.

Take the startup campaign, 5880 Clicks / 716007 Impressions = 0.82% CTR.

That is a pretty poor CTR and explains the high cost of website clicks. If the CTR was up around 2% the cost of website clicks would be halved. Would probably look at changing the ad content if CTR was this low.

It is low but not unusual for Facebook.

I advertise a few mobile apps on facebook and have an average CTR of 2.5%.

Point is if your ads are getting that low a CTR than you need to change the content, audience etc. No point in running ads for website clicks when no one is clicking on them.

Pardon me if this is off topic, I'm sure someone will let me know. I run an ad on Facebook for $25 for the weekend. We get somewhere between 2000-6000 "reaches". We get a fair share that like the post, I "invite" them to like the page, hardly any do... Is this normal?

Liking the page would require effort from the user and most are reluctant to deviate from their current thought process.

Most people are on facebook to scroll down their news feed and maybe like a post or 2. There is no motivation for them to go to your page and just 'like' it, unless you have content that appeals to them.

People are bombarded with advertisement all day every day, they simply feel no obligation to help you unless you can provide them with something of value.

Think about from your own perspective, how you use something like facebook. Do you go and like random pages from people sending you unsolicited invite requests? Probably not, so why would anyone else do it.

I don't disagree with anything you've said... but... Since they already liked the post, I'm not picking random people on Facebook, I'm asking them to repeat what they just did. I think it comes to them just like a friend request.

I know it's not random, but liking a post does not mean a person is obliged to perform a further action for your benefit.

My understanding is an invite will just push a notification to a user saying 'Such and such invites you to ...'. Its different to a friend request as that has its own context where you accept, reject. People get notifications all day, so it's easy to understand why they might brush over an invite for some other content.

A little side note, someone liking your page doesn't mean they are going to see your content. Only 10%-15% of people who like your page will actually see your posts on their news feed. Have a read about it on the net, thats a facebook policy that's been in place for about 4 years now. If you want to extend your 'reach' you will need to boost your posts for a fee.

Sorry reply might come across as bit negative, I just wish to give you some realistic expectations. Keep doing what you're doing, keep promoting, keep growing.

I didn't take it as negative at all. Anything that moves the ball forward is good. Thanks for taking the time.

I'm looking for the next step to take when someone has liked a post, and inviting them to like the page seems to be the only option, which as I said, has limited results.

Your comments are helpful.

That spend is too tiny to produce anything useful and reach is probably the wrong metric for you at that stage. Focus on onsite conversions. Likes might be helpful directionally but don't pay the bills.

Sounds like it's best for mindshare/brand-awareness rather than action beyond that?

I think facebook works for people that sell products that are revelant to a person's friends. Even if you need a tool to store local files encrypted do you college, or high school friends need this? If not you probably wont share it and the viral effect will not happen.

This is a superficial complaint, but please fix the misuse of decimal points and commas. They are distracting from what is an otherwise interesting read.

It's actually quite common in Europe to use commas for decimal points, and periods for thousands delimiters.

I learned something today.

I wonder if you would have done better using native advertising through key people in your industry who are followed by possibly customers.

I've been out of the loop on Facebook ads for 5 years now. Is there a good conversion tracking story now, or no?

Yes, the pixel code is pretty decent. You can set up custom event tracking in JS on form submit or whatever you like, and their algos optimize for that reasonably well. It's still more about audience and ad design than magical pixie dust, but it seems to work.

Perhaps they could improve their offer, ads, funnel, landing page, image, or a variety of other things..

Yep. The whole user flow for a Facebook lead might need to be different. I try and treat acquisition channels as differently as possible, within reason. For example, your organic traffic may behave completely differently than your Facebook Ads traffic and that may be happening for a variety of reasons. Organic traffic is targeted and active. Someone was likely searching for something highly related to your product. The same goes for paid search traffic. However, Facebook and other display ad traffic is passive. You're targeting someone who may be interested in your product but they aren't actively seeking it out. Those clicks are much different! I'd probably throw a much heavier sell at those people and make sure your actively on-boarding and engaging them to use the product throughout the trial period. They are likely to be far less motivated than other types of traffic, so it's your job to get them motivated!

"Facebook may just be the wrong audience for a B2B tool."

Duh.

Thanks for sharing, informative article.

Derek Muller burned money on Facebook ads, and did an analysis showing that their metrics are flat-out horseshit. Definitely have a watch if you're considering spending on Facebook.

https://www.youtube.com/watch?v=l9ZqXlHl65g

How do you spend 20k on marketing without testing conversions with a smaller budget first? They could have spent 300, noticed it didn't bring in anywhere close to 300 in sales, and aborted.

How I picked 300:

They're charging 20-80/mo for the product. Assuming they're willing to pay 100 to acquire a user that hangs around for a while, by the time they spent 300 they'd know if they picked up at least 3 users or so. Less than 300 isn't enough to be significant, more than that is wasteful.

That doesn't really work on pay-per-click type models. You need to be acquiring 100+ customers before you can really measure properly.

You can't measure "properly". You can measure enough to know that you'd be blowing it if you spent 20k.

If spending 300 didn't bring in at least 3 users over a week, that advertising was never going to work. If it did, then you go up an order of magnitude and test again. Never scale more than one order of magnitude at a time.

You really don't know that. It takes time for the algorithms of companies like Facebook to optimise properly. For my company, CPA has come down significantly as we have spent more.

Reply on news.ycombinator.com