Using Data Analytics To Fix A Broken Acquisition

By the end of Year 3, it was abundantly clear our first acquisition was in trouble.

My Spouse: Uh, That’s Our Retirement Money…

I married an accountant, by the way. She’s had a full decade of experience debunking my boundless optimism. So when we wrote a four figure check for the deal, there was a firm earnings target on the table. No hand waving allowed, how much will we make per year?

Which we had consistently missed by over 50%. A fact which was brought up every time I asked to fund anything else….

So… What Went “Boom”

When a website blows up after being purchased by a new owner, the usual suspect is losing your traffic and Google rankings. The good news is that this particular horseman of the apocalypse was held completely at bay. Look at this trend in average monthly visits…

Website Traffic – Retained & Grew Audience!

I mean, look at that! Such Rankings! Much Traffic! So SEO!

It’s the Revenue, Stupid…

Which means this entire miss was on my shoulders. A bad revenue forecast.

The previous owner had not monetized the site, so we had to build our own forecast. We had another site in a similar niche, so we weren’t betting totally blind. That website used a mix of display advertising and affiliate offers. What we missed was that while the other site was technically about coding, the underlying demographics were very different. Display ads earned significantly less on the new site.

Once we figured this out, we now faced a double barreled problem:

  • Not enough traffic to make display ads really work
  • No “high value” offer or ads to boost our returns

While we had priced the deal at a 30% return (25% after expenses), this issue kicked our cash-on-cash returns down to savings bond levels. Between having revenue cut in half and offer testing expenses, our return was in the single digits.

The Solution – Build Traffic & Get Better Ads

At this point, I realized how badly I misjudged the audience. It had become apparent that our traffic was mostly graduate students instead of developers. The latter had money and reason to spend it; the former was an inscrutable mystery to me.

It was time to rethink things. Successful business turnarounds usually involve playing to your strengths rather than fixing your weaknesses. Could the same principle apply here?

The first pass at the analytics suggested our (well established) SEO playbook could work in this niche. Adding more content to grow website traffic was an option. While the traffic wasn’t going to earn at the same level as our other site, it would help cover the gap to plan. More importantly, it could qualify the site for a better advertising network and bump our rates up.

The affiliate strategy wasn’t dead either. The more traffic we had moving through the site, the easier it would be to test different offers and collect feedback. But first things first. We put the affiliate program on hold and buckled down on content creation.

Content Development: Stacking The Deck

Our vision was to add several substantial sections to the site, blanketing a selected topic with articles to build authority. We allocated about 40 hours to the effort, split across a couple “content clusters” of related articles. This took us about a month (part time gig).

The content clusters were selected using data science and common sense. We used two analytical tools to identify topics with weak content. This was validated by inspecting the search results: if it looked lousy, we were in business! Once a target topic was selected, however, we attempted to make our coverage as comprehensive as possible (even if a specific article was unlikely to rank).

The best example of that was in the first wave where we basically carpet-bombed a topic. As you can see, it worked…

New Articles Significantly Outperform Existing Content

The impact of these analytics on our overall return on content was significant, resulted in articles which significantly outperformed the original owner’s content. Our worst clusters of content generated more than five times as many visitors as the articles which were already on the website. Our best content clusters did even better. The two keyword research tools that we used are profiled here (free accounts available, see recommended tools).

Seasoned Domain, Faster Payback

While keyword targeting clearly made a difference, the project benefited from the acquired website’s strong authority. Google took notice of the new content within a few months and consistently ranked these articles higher as we moved into the following year.

This is a rough guess, but deploying this content from an established domain likely accelerated this process by six to twelve months. It’s a fast path out of the Google sandbox. We’ve done similar projects with new domains and had slower progress.

Once we started getting traction, we repeated the effort. We’ve launched a total of seven additional waves of articles for the website, six internally developed and one (larger) that produced by outsourced writers. Higher success rates make the economics attractive.

The Jump to LightSpeed: Ezoic

I breathed a sigh of relief once our traffic crawled north of 25,000 unique visitors per month. It was party time. We could start to shop our business to higher ad networks.

We flipped the website over to Ezoic’s platform, which helped us at three levels:

  • Automated optimization of website layout for specific types of users
  • Higher advertising rates due to more competition in the auction
  • Sped up key pages using a free CDN and file compression

The combined effect of these improvements resulted in a significant increase in advertising revenue, which accelerated over the next several months as we made broader use of their capabilities. As the chart below demonstrates, the shift to Ezoic can be credited with tripling our overall advertising revenue run rate. We just signed up for their Premium program (direct sales placements), which we expect to drive further improvement.

The Ezoic CDN turned out to be a major win as we grew our traffic. Most WordPress sites are not terribly efficient. Moving any incoming traffic over to the CDN reduced the load on our infrastructure and allowed us to scale without adding servers or re-configuring the site. So a big win with Ezoic, at multiple levels (see our recommended tools section for a link to a free 30 day unlimited use trial).

Paths Forward

What a turnaround!

Between the increase in traffic and improved advertising performance, the average monthly revenue for the website has increased by 44 X. We’ve gone from being 55% below plan to being on track to deliver a 635% cash-on-cash return.

The journey is even more dramatic if you factor in the likely price we could sell the site for. For our final chart, here’s a “net profit on flip” view which calculates the net profit at every point since we launched the new strategy. This is “all inclusive”: purchase price, migration cost, content costs, revenue earned to date, etc.

While we have no intention of selling (not done yet!), our projected return from the deal has swung from a 40% loss to a 12 X gain on our initial investment, under conservative assumptions (3 month run rate for a fast growing site). With plenty of free cash flow to replicate our growth strategy across similar niches…

Fair Market Value for the Property Continues To Grow

The ultimate validation? My wife asked who we should buy next…..

Leave a Reply

Your email address will not be published. Required fields are marked *