It’s Episode 5 of The Numbers Game and this week we're tying up the background story.
The first segment covers the last six months of The Podcast Host, bringing you right up to date on things going on in the company – focusing more on the software development side of things.
It’s the turn of income stats in analysing our KPIs this week, while our lesson looks at the pros and cons of working with agencies compared to hiring your own staff.
In our experiments, we’re continuing with the theme of email optin rates, and finally your homework – try out a few more A/B testing variations for your leadmagnets. Enjoy!
Story | Developing SAAS & Making Content Stacking Easy
At this point our client work wasn't scalable (trading time for money in producing audio content, week in, week out) and it was holding back the business. We wanted to do what we enjoy, which is creating great content ourselves. That's where our refocusing comes into play.
We moved away from clients (with the exception of a few projects) and instead to the idea of allowing people to create high value content with a single app.
It made me think, can we craft an app that makes it super simple for a podcaster to produce their episode?
From this we added content stacking – the idea that you could craft one content idea that turns into a variety of forms. For example, a video that could be published on YouTube, the audio track on a podcast, and a summary published as a blog post. One content idea, three bits of content. Simple.
After receiving SMART funding (which was touched on in last week’s episode) the project met its first barrier – employing an agency.
We looked at universities, but the amount of bureaucracy made it difficult to get things done. Next we tried some agencies, both big and small, but the large fees or lengthy timescales didn’t work out.
So, where are we now? We’ve decided to go it alone. Two months ago we hired a back-end developer to work part-time for us, building a fast and rough prototype. Now, we’re on the lookout for a front-end developer to help us build a slick interface for the app.
While there is a bigger risk and not as much quality control as there would be with an agency, we believe it was worth sacrificing to build our own team – even if it takes a little longer.
The plan is to get a Beta out during the summer, and we’ll be looking to recruit some of our existing audience to try it out and see what everyone thinks.
So, that's the back story. We're now up to date! From now on, the story segment of the show will concentrate on the weekly goings-on at The Podcast Host – a behind the scenes of the ups and downs in the team. Looking forward to it!
Data | Where the Money’s Made: Clients, Membership & Affiliates
It’s week two of looking at the Key Performance Indicators (KPIs) at The Podcast Host – and this week it’s the turn of our income stats.
As mentioned above, we still have some clients and some bigger one-off projects. For these projects we only really take them on if they’re above the £2000 ($3000) mark. Our standard client work is above £400 a month, which is a good regular, recurring income.
In terms of membership, which is growing steadily, we are adding a couple of people per week on average. That takes us up to around £350 per month. We have 25 members right now, and the hope is to grow that to around 200 members by the end of the year.
Next up, affiliate income. This is the main source of income here at The Podcast Host. There has been a slight drop this month. We normally receive around $100 per day in affiliate income, but one day around 3 weeks ago it dropped down to around $20, then the next day it was $25 and the next day it was $15.
Affiliate income is so unreliable for a number of reasons and that was worrying for us. Thankfully, we’re back up to similar levels as before, and for March we received around $4000 in affiliate income.
That's the report for this week. On Episode 6 of The Numbers Game we’ll be looking deeper into the data behind our email marketing.
Lessons | Agencies or Employees? Take the Risk & Build In-House Capacity
The main lesson from the story this week was related to the choice between agencies and employees.
The reason we wanted to go with an agency was because we’ve never done something like this before. We’ve never worked with software development and didn’t know if we would find the right person or be able to manage the process of work.
So that was the plan. We’d work with an agency for six months or so and learn from them – and make sure it was of a really good quality. But in reality this was just a case of being afraid to go it alone.
Over that time we talked to around five or six agencies whose quality varied massively. They weren’t fully invested in the project going forward or as fool-proof as we originally thought. That’s why we decided it was worth the risk of just hiring our own employee.
So that’s the lesson:
If you want to build a piece of software long-term then realistically you need staff. You need to jump in sooner or later.
Experiment | SumoMe’s Upgrade Issues & Leadmagnet Results
We’re back continuing on with the email optins work that we started five weeks ago.
So far we’ve designed our own custom leadmagnet, implemented it on the site, signed up to an email provider, created a welcome email sequence and then finally we talked to you about the software itself.
Now, it’s time for a rant. SumoMe. Let me qualify by saying that I love the company that Noah Kagan's built, from the personality to the product quality. I've been massively happy with their List Builder tool up until now. But recently they upgraded the tool, giving it a complete overhaul. Fair enough, you have to upgrade software from time to time, but the upgrade was handled really badly in this case. I went through the process only to find that a lot of my popups were lost entirely, ALL of my campaign data was lost and that every popup now needs updated into the new format if I want to use and iterate on it in future.
They’ve changed it from the old method of A/B testing which was campaigns, to ‘tests’ – whereby you get a popup and variations of that popup. I've got my head around it now, but I lost nearly two week's worth of optins because of a complete lack of documentation. I was searching around like crazy looking for help, but all the docs were for the old version. Who releases a new version without ANY kind of support docs?
Eventually, after a number of emails (and the help rep was really helpful, I'll give him that, but of course it took days of back and forth to go through all the problems) we were able to look back at our stats…finally.
In previous weeks we changed the generic ebook across the entire site to give us a baseline for our optins.
On the equipment pages it was bringing in a 1.4% optin rate, which was fairly low. On the Skype pages we had a 1.2% optin and on the fiction pages we saw a 0.7% rate.
Then, on all our remaining articles, it has a 2.3% rate – which we are seeing as our baseline optin rate.
Our equipment pages are the most popular by far, so the aim is to get a leadmagnet for those pages that will improve on the 1.4% rate.
Now, looking at the individual leadmagnets for their own sections:
- The ‘Ultimate Kit Bag PDF’ had an optin rate of 0.4% on the equipment pages.
- The Skype Interview Checklist had an optin rate of 1.5% on the Skype pages.
- The ‘How to Create a Fiction Podcast’ book was bringing in a 1.4% option on the fiction articles.
So, a slight negative at the end of that experiment run – custom leadmagnets really aren't working for us! If we were just testing one custom leadmagnet, I'd wonder if we were doing something wrong ourselves… But no, this was three magnets, all in different parts of the site. It's a technique that general wisdom seems to claim works really well, but we found, so far, that it's really not for our audience. Hey, that’s the whole point of experimenting!
What do we recommend? Try create one custom leadmagnet for your most popular article and test it out. It might well work for you, but don't invest a whole lot of time in it until you know.
Actions | A/B Testing: Find the Right Approach by Testing Testing Testing
It’s homework time. To tie up, even if you are doing generic leadmagnets, then A/B testing is well worth it.
So, you’ve already created an A/B and a couple of variations of headlines to give you an insight on how different they can be. Now, we want to continue this a bit more.
Try a few things:
- Test out the difference between having an image or no image. We’ve found that some popups perform much better without an image, and others with.
- Try different headlines. Find four different variations and mix it up as much as you can.
- Vary the length of text. It’ll be audience dependent, but changing the body text from short to long and vice-versa might make a big difference. It does for us – long text performs much better.
Try out as many variations as you can, then narrow it down to find out exactly what works best for you.
What do you think?
We’d love to hear from you, so let us know how you got on, what you think we’re doing well or anything we can be doing better!
Or, we’d appreciate it if you could drop us a review on iTunes.
Other Resources mentioned: