Domain Performance Analysis

Domain Performance Analysis

Understanding performance often requires getting really deep into the numbers. The other day I said to a client, “We like to understand which domains we win and which domains we lose versus a baseline.” Let me take you through some high level data that gives a snapshot of what I’m talking about.

For a start the data I’m about to go through is actual numbers for a client but for obvious reasons any identifiable information has been removed. The data is for a particular domain portfolio where it would appear that our optimisation services were struggling.

Escrow

The portfolio being tested has a baseline of around $95.59 per day and within the first week of being tested we underperformed by $12.37/day. Compared to what we usually experience this was a disappointment for ourselves and more particularly for the client. So what was going on?

Here is a snapshot of the performance data from 4 weeks, 3 weeks and the most recent week. All data has been normalised to reflecting the average daily for the period. There is the current revenue, baseline revenue and then Daily performance and the percentage difference.

Analysis

The columns are in bands of traffic. From All domains to those doing more than 0%, 20%, 40% etc. up to greater than 100% of the traffic compared to the baseline.

As an example, four weeks ago for domains that we saw at least the same amount of traffic as the baseline we performed at 100% of the baseline….in other words, no uplift. Overall, we performed 87% of the baseline for ALL the domains…..not good.

Most domain investors would immediately throw up their arms in horror and bail on the test at this point…..but let’s take a look at what happened over time.

I’m going to jump right down to the latest series of data for the past week and compare this to the baseline. For domains where we saw at least as much traffic as the baseline we are now performing 15% better. In fact, ParkLogic performed better for domains that have at least 60% or more of the baseline traffic. The overall performance jumped from 87% of the baseline to 96% of the baseline in a few short weeks.

What this data really says is that if there is traffic we perform better…..if not, then we don’t. This kind of makes sense! All that we really needed was for time to go by and those domains with little traffic generate a click or two.

To see the change in performance over my data set I summed the % Difference rows. This is a crude way to find out overall what was happening each week. The summed percentage moved from 634% to 711% which is an overall increase of 12% and to date this is continuing.

This is a lot of interesting data but what does it really mean? The problem with the portfolio isn’t that we aren’t performing, it was that it was taking a bit longer than normal because the typical domain has such low traffic. Unless there is traffic we really can’t quickly do the optimisation to increase the result.

What’s important is that as time goes by the trend is continuing upwards as our algorithms and processes go to work on the portfolio. What was also hidden in the data was the fact that we are currently providing an additional $40/day in revenue over the baseline for a large portion of the domains. In fact, if we were to keep managing all the domains that we were winning on, handed back the domains that we weren’t to the baseline company (assuming is performed the same as it previously did) then the client would receive a net revenue increase of 42%! That is a HUGE number!

The overall current number of $3.71 down was actually hiding a HUGE windfall….the challenge is to work out how to get at it. We are continuing to monitor the results each day and in our meeting with the client we plan on working out the next steps to extracting the traffic value.

There are two parts to this analysis which are key. First of all, you need the ability to mobilise the data and secondly, an eye to interpret what the data is actually saying. I’m not meaning to blow my company’s horn but the ParkLogic team has spent many years developing processes and skills to do exactly these two things.

---------------------------------------------------

Michael Gilmour has been in business for over 32 years and has both a BSC in Electronics and Computer Science and an MBA. He was the former vice-chairman of the Internet Industry Association in Australia and is in demand as a speaker at Internet conferences the world over. He has also recently published his first science fiction book, Battleframe.

Michael is passionate about working with online entrepreneurs to help them navigate their new ventures around the many pitfalls that all businesses face. Due to demands on his time, Michael may be contacted by clicking here for limited consulting assignments.

Tags:
1960 Hits