Thirty months of Paperight metrics

Here’s a series of graphs about Paperight’s journey. They cover two and a half years, by month. We’ve picked our starting date for these as September 2011, when our investment and funding from the Shuttleworth Foundation kicked in. I’d already been working on my own, part-time, on the model and prototype tech for almost three years before that. We’re including:

  • our high-level projects and decisions
  • our team members and size
  • sales by quantity
  • sales revenue in US dollars (before deducting rightsholder earnings)
  • social media followers
  • unique visitors to

What’s interesting here is how little correlation there is between any of these. Perhaps our small data set lets outliers hide any evidence of correlations or trends over time. Or perhaps disruptive startups are just messy and unpredictable. (One very recent stat these graphs don’t show is that in the first two weeks of June we had our highest ever site visits, and our lowest ever sales. Go figure.)

This first graph shows our high-level projects and decisions, the most significant being the realisation in February 2014 that our original business model doesn’t work, and we need a big pivot. (More about that in this post.)

Click on each graph to open it at full-size.

High level projects and decisions

If you use the monthly-archive menu on this site, you’ll find posts by team members from the months concerned.

Then our team: we grew our team quickly from about five months in. This was critical in getting our outlet footprint established and creating our own content (reformatted classics and grade-12 past-exam packs).

Team members showing team size

So how many books did we get out there? Here’s books distributed by quantity. That is, the number of books for which our outlets paid licenses and printed. There are three big spikes that are outliers: large bulk deals we initiated and closed working directly with schools and sponsors. These look lucrative, but the cost of making each sale was very high, so they were not necessarily profitable.

With hindsight, I would have liked to have measured the time our staff spent on each sale. It would range from zero for organic sales to dozens of hours for the large bulk sales. At about R100/hour for staff time (cost to company), even two full days on a bulk sale could wipe out our margin. Often they took more than that: finding beneficiary schools for a sponsor, analysing what students needed, meetings, handovers, and more.

So by late 2013 I did not push our team to seek out bulk sales and focused rather on strategies (like the #textbookrevolution campaign, see posts by Marie, Tarryn and me) to grow organic sales, which I believed would make for a healthier business in the long run.

Now compare revenue.


Let me take you through that:

  • For our first nine months (from May 2012 launch), most books on were free. That is, we charged no licence fee for our grade-12 past-exam packs or our classics. And at this stage we had very few books from commercial publishers in our catalogue. Downloads are satisfying (we didn’t know then what ‘high’ or ‘low’ would look like), but revenue is almost nil. We’re getting customers but no money.
  • We decide from January 2013 to start charging a licence fee (usually 1 to 2 US dollars) for each copy. Revenue goes up a little, albeit in bumpy fashion. But downloads almost vanish. That is, most of our users disappear, but we do, technically, make more money from those that remain. With hindsight, I want to think we started charging too soon. But I also know that if we’d waited, we may only have postponed discovering the sad truth about our original business model.
  • Faced with these now disappointing sales, we decide to go find some big customers. We send Yazeed on a sales course and he throws his time into finding sponsors to buy for schools in bulk. We get a few bites and by the second quarter of 2013 we think we’re onto something. I’m not yet thinking about what each sale is costing us in time, only that bulk sales may be the way we bring in cash while we grow. But our first few deals come through existing relationships, and we don’t up our game building new relationships. So we run out of good leads. By the third quarter of 2013 I realise how much it’s costing us in time, and therefore payroll, to find bulk sales. And I see that beneath the tall trees that are the bulk sales, there’s no mulch. No organic sales.
  • By early 2014 our sales look a little better. Perhaps our smarter marketing work is paying off, and we’re also finally selling popular study guides from major South African publishers. But it’s too little, too late for our original business model.

On to some web metrics. I don’t believe Paperight has ever been an online business. We’re part of the offline paper-book industry, and we just use the Internet to scale our footprint. We only realised this about ourselves in late 2012.

Nonetheless, I expected to see some correlation month to month between site visits and sales. However, we’ve never noticed any real correlation we can count on or plan around.

Unique visitors to

No metrics are more vain than Facebook likes and Twitter followers:

Facebook Likes and Twitter followers

We are hugely grateful for the moral support of thousands of people over the last two years, represented well in this graph of social-media supporters. And insofar as Paperight’s mission was to change the way people think about rights and book distribution, perhaps these followers represent success.

But as you can see, they have zero correlation with sales. I know that for our next phase, our big pivot, I won’t spend a heartbeat’s time trying to grow them.

That said, please keep supporting us anyway. It keeps our heads up when things are tough. We have a long way to go and much more to learn. We’ll keep sharing.

One thought on “Thirty months of Paperight metrics”

Comments are closed.