IRL Experience Design

Everything Else

Digging Into Online Stats to Identify Issues

The 10 days of SXSW were pretty dismal for the Eureka Room. But before I made the assumption that it was due to SXSW, I thought it would be a good time to dig into the stats and make sure there weren’t other things going on.

Customers tend to come from online searches, word-of-mouth, and drive-bys. A typical online search might be: Google “weird austin things to do” > click on the do512 result > click on eureka room > go through the buying process.

I can lose a prospect anywhere in that chain. When sales go down, it’s really tempting think you know what is causing it (usually something that is fresh in your mind), but let’s not fall for the recency bias and instead look at the data.

Here’s the strategy I used.

Calculate Traffic and Sales

  1. Get a benchmark period. If you think sales have dropped, determine an earlier period that you think was “normal” and note those dates.
  2. Determine the “low” period dates.
  3. Calculate the average daily visitors for each of those period and the % drop. This alone might be insightful.
  4. Run a report in bookeo to see the weekly

Possible Causes For Drop in Traffic

If the traffic has dropped “too much”, then it could be due to one of these:

  1. Temporary (SXSW) competition or new permanent (similar business has opened) competition.
  2. The weather affecting people’s choices. (Good weather = fewer people want to be indoors).
  3. Referring sites not referring as many people (due to a change on their site or a change in their rankings)
  4. Natural search results for eureka room has decreased.
  5. A decrease in google ad (or other ad) spending means fewer ads driving traffic.
  6. Less word-of-mouth and driveby effects.

This is NOT an exhaustive list. There are many reasons traffic could drop.

Possible Causes For Drop in Sales

If traffic has NOT dropped or if it has dropped but sales have dropped even MORE, then it could be due to one of these:

  1. A decrease in qualified candidates. It’s not just keeping up good traffic, but keeping up good traffic.
  2. A decrease in eureka room booking availability. (e.g., for a while we cut back on hours because we didn’t have the staff which led to less bookings)
  3. Something has changed on the eureka room website. Maybe the copy was changed to something less compelling. Maybe the booking path got longer or more complicated. Maybe the site doesn’t load well for some users.
  4. We have some recent bad reviews turning people off.

Again, this is not an exhaustive list.

My Diagnosis

I choose these periods: Jan 16-Feb 16 and Feb 16 – Mar 16
I chose these because the first week or so of January was extremely high. Also I changed the website copy around 2/15 (I think).

Comparing these two periods, we have a 12% drop in web visitors and a 19% drop in sales.

Important: a “sale” can be the day the booking was made OR the date booked.

  1. Temporary competitions: SXSW. This coincides very heavily with the drop in sales but does NOT coincide with drop in traffic. It was pretty much constant.
  2. Good weather. Perhaps, but it was also good weather in the benchmark period.
  3. Referring sites: All sites performed similarly. One BIG drop was the “direct” source. The “direct” is the catchall analytics uses when it doesn’t know where stuff comes from.
  4. Natural search results are hard to know. I believe google has a tool for this, but I don’t know.
  5. Google ads are hard to evaluate. Maybe they got smarter and better targeted so less traffic but higher sales (which is the real goal). Turning them off did not seem to affect the traffic, but it did effect sales (see below). So it’s possible the traffic I lost from turning off ads was made up for by another source.
  6. Word-of-mouth and drivebys are hard to evaluate. We could get these numbers from the bookeo field “how did you hear about us”, but since it is self-reported, I wouldn’t trust it too much.

On the sales side:

  1. I had a decrease in ad-spend on google ads. I stopped all google ads on 1/25. Comparing the period 1/8 to 2/4 (when ads running) to the few weeks after (but before SXSW), Google ads claims that we got $1001 in sales from that $368 spend. Remember that many apple products block tracking. And also that most people don’t book right away since they need a group. They probably send the link off to friends and maybe someone else books it or at least buys their own ticket(s). Unfortunately due to some analytics issues I had the “google ads tracking” in bookeo turned off so I don’t know if any of the ads I ran to jumpstart SXSW worked. But I think if I can get $1001 from $368, that’s a pretty good return for now. That said, I hate google ads.
  2. We definitely had more limited availability. I closed down Wednesdays and we also didn’t do all the hours we could have on the weekends.
  3. The copy on the website was changed around 2/15. This does not seem to have made a difference. Though I only have a couple weeks before the effect of SXSW starts to muddy the numbers.
  4. Overall the reviews are good with about a 4.8 rating so I don’t think this is a factor.

Overall Conclusions

  1. There is a strong correlation between website visits and ticket sales. So more traffic will mean more sales. More ad spend is needed: posters, online, events, etc.
  2. Google ads are probably worth starting up again, at least until we start to get really booked up.
  3. Website copy change is something to keep an eye on, but does not appear to be hurting.
  4. SXSW was likely the biggest cause of the sales drop.