After blogging about the start of Write500, I told myself I’d write a follow-up when I reached 500 subscribers. That happened a lot faster than I expected it would.
Yes, I’m at 500 opt-ins – barely 10 days into this.
In this post – expect stats, charts, breakdowns, and guesses as to what sort of marketing worked for me, what did not, and where I plan on spending my time next.
First, I need to explain the terms I’m going to use in this post.
- Signup: User has entered their email on the subscribe page
- Opt-in: User has clicked the “Yes, I’m in!” button in the confirmation email
- Opt-out: After having subscribed, the user logged in and disabled the prompts
- Recovery: The user Signed Up but did not Opt-in within 3 days
A note about Recovery: When I first launched this, I had the system simply delete their email address if there was no signup within 3 days. The first time it ran, it knocked off almost 10% of my overall list.
I then decided a better idea would be to try sending out automated reminders. So I built a Recovery system – after 3 days, instead of being deleted, the user is moved out of the subscription system completely (they “vanish” as a Signup, since I can no longer reasonably assume they’ll opt-in).
Once they land in Recovery, three reminder emails are sent (with different subject lines, every three days), and if they still haven’t opted in by the third email, I take it as a loss and remove them from the system entirely.
Right now I’m not tracking the amount of users I’ve Recovered this way, but so far I’d say the percentage is around 10% or so.
My First 500
Now that we have that cleared up, I’m going to show you this chart:
As of the date of this chart, I have 560 Signups. The light blue on Jan 9 there is the BI tool (Domo) predicting how many new subscribers I’m going to end the day with. I’ll get into the reasons behind that growth later in the post.
Now, for the rest of the context, I’ll show you these two:
Of all the Signups in my database, 88% are active (exactly 500 at the time of this screenshot). The green wedge there is Pending, and the little orange wedge is Opted-out.
I also have 61 users currently in Recovery. 75% of them have been attempted once (meaning they’ve received one email asking them to confirm), and 25% have been attempted twice.
Now let’s work out where they came from.
Social is the winner at 56.9%. Direct, annoyingly, is the second-largest, and is most likely also the result of social. If users try launching my site directly from their mobile apps, the analytics tag counts that as Direct, since they appear to land on the page out of nowhere.
Social + Direct accounts for 91.5% of the traffic to my site. Which should not be surprising, considering Write500 is brand new, and I’m running two ad campaigns on social networks.
Also overwhelmingly: mobile.
I’m expecting that to be a direct result of the fact that most of my ad spend so far has been on mobile.
In terms of organic sources (content, inbound and search), I’ve had less than 10 signups so far. Which is expected – organic takes a very long time to build up. Easily 95%+ of the traffic I’ve received so far, is paid.
Paid Advertising Results
And now, we get to the heart of the matter – my adspend results.
I ran campaigns across a few sources:
- Facebook – Promoted my page
- Instagram – Specifically drove people to my site
- Project Wonderful – A niche display network mainly serving creative blogs
- WritersCafe.org – The only writer forum that enabled me to advertise
I’ll start with the niche networks
For my $25 in ad spend (which was doubled to $50 since it was my first purchase), the site claims to have driven almost 60’000 ad impressions across 2 creatives, but resulted in less than 20 clicks. That’s a CTR of 0.03% – dismal even by industry benchmarks.
Those 20 clicks went on to drive 2 signups, for an effective out-of-pocket CPA of $12.50/signup – making them the most expensive users I’ve ever acquired.
In short, I’m not advertising here again.
Network: Project Wonderful
I first came across PW many years ago, as far back as 2008. They run an auction-based display ad service, and mainly have blogs in their network. I decided to commit $50 upfront, and start with very low daily spend targets.
Marginally better performance than WritersCafe. I get a CPC of $0.30 on average, but an even worse CTR of 0.02% (with 27 clicks off ~135000 impressions). In total, I’ve only spent $8.02 of my original $50 budget.
There’s a good chance I could improve my CTR by updating my creatives – right now they’re mainly just the Write500 logo on a blue background, and visually, I’m competing with much more lively, colorful ads.
Unfortunately, despite sending all that traffic to my site, I have not seen a single signup originating from Project Wonderful.
When I have time, I’m going to try optimizing a lot more. I have a few landing page and creative ideas to try, and I can opt to bid for full-day slot presence (not just rotating to try hitting a target CPM).
For now though, my effective CPA is NaN/signup.
This is where I’ve received most of my traffic. I’ve run two campaigns – this one was a broad promotion of my site.
Those are solid results. That’s an acceptable CPC, and the CTR was improving over time. For a total spend of R700, I drove 649 clicks to my site, for a CTR of 1.48% (which is awesome) and an average CPC of R1.08.
Unfortunately, I launched this campaign before I had the Facebook pixel installed, so there’s no way to attribute any signups to it – but just looking at the overall traffic sources v signups, I’m convinced that Facebook has driven a large % of my users.
I’m definitely doing this again – I’ll be experimenting with different creatives next time.
Other metrics of note – I let Facebook decide on placements automatically (which is the default), and it ended up generating much more traffic via mobile. These are how the clicks break down:
Another piece of curious info to note – where the audience ended up coming from:
This was honestly the best result of all the advertising I’ve done so far, because it solidly indicated two things: Most of my users will be on Instagram, and most of them will be younger women.
That’s the cheapest market research I’ve ever done. R700 for a data-driven insight like that is a bargain!
Based on that insight, I then launched my first “proper” marketing campaign – targeting Instagram only, filtering for younger women, with 3 different creatives.
This has been the single largest driver of signups to Write500.
I had the Facebook pixel set up for this one, so I was able to track conversions accurately this time. The campaign has only been running for 4 days (today is the fifth). In that time, I’ve spent R378.69, and generated 232 signups, at an average CPA of R1.63/signup.
This was a very specifically-targeted campaign. I went for Instagram, in major English countries (US, UK, ZA, AU), with age range 16-22.
Most of my conversions came off one creative (pictured above), but I believe that’s a budget restriction. With more spend, Facebook would run all 3 creatives. Right now there’s not enough data to draw any conclusions about relative creative performance.
That’s not all, though! When I set up the campaign, I linked it to the Instagram account for Write500, and I had already posted 2 images to it when the campaign went live. I’ve planned to post one image per day, for the rest of the year – all formatted quotes like the one I used in the ad.
Not only has the ad itself gathered almost 2500 likes (as of this writing), I’ve also had a burst of new followers.
This was a totally unexpected side effect of the ad campaign. Even after I stop spending on instagram, my daily posts will now continue to gain organic traction.
The headline conclusion: Mobile wins.
For Write500, I had a mobile-friendly signup page, and so far I’ve received most of my traffic and signups from mobile sources. Going forward, if I build new sites, products or projects that I intend to monetize, I’m going to be building mobile-first.
The corollary conclusion: Desktop display advertising is dead. Maybe not dead dead, not yet, but between already-low CTRs and the rise of adblockers, I suspect the effectiveness of desktop-only display now ranks the lowest.
I have not yet tried paid search, though – that might be the exception, and I’ll give that a go at some point.
- Small niche networks are a waste on a limited budget – rather spend on big networks.
- If you’re targeting the younger generation, Instagram in particular is fantastic.
- Data-wise, it’s definitely worth setting up the vendor tracking pixels on your site upfront.
- Attribution is hard enough already, so it’s worth taking extra time to build tagged URLs.
My Instagram campaign will end in a few days, and I expect my subscriber acquisition rate to drop to near-zero. Which will be fine: I’ll have over 500 opt-ins by that point, and I’d rather go back to spending more time working on the product.
The biggest thing I’m working on right now is a mobile app. Pretty basic stuff: Users will be able to authenticate, manage their settings, and get push notices to remind them of daily prompts (and to write their 500 words).
That app will be free, and I expect my next block of adspend will be on promoting that app. Before it even hits the app stores, I’m going to thoroughly tag it for analytics.
Beyond that, my next biggest thing is going to be finishing the annual content pipeline (I only have 33% of the daily prompts loaded so far), so that I can stop stressing about that. And then starting work on the paid version of Write500 – which is how I intend on eventually making all of my money back.
I hope you enjoyed this post! I’ll be writing more of these, as I learn new things on this journey – you can subscribe to my blog using the Subscribe widget on the very top right.