I’ve never had much luck in the past finding good statistics against which to benchmark an email marketing campaign, and I think that one of the reasons is that often it is difficult to draw a direct link between marketing message and subsequent action, unless of course that action is something like a click through. When I started working with Air Pacific in Fiji on an effort to increase the number of hotel bookings they made online, I was far from an expert on email marketing, but I knew one thing; that was that I wanted success or failure measured by increased bookings (ie. sales), and therefore increased commission revenue earnt by the airline from my efforts. I was not prepared to settle for success based on open rates or click through rates.
Air Pacific had a one year track record of selling hotels, using a white label (private label) solution from Expedia. Immediately, we managed to triple the number of hotel sales made on their website with all else remaining unchanged, so the email marketing is the only factor that can explain this growth. This was a massive success, especially considering only half of the passengers were actually sent the email in question. Latest numbers I have seen show the increase getting closer to four times the average monthly sales prior to our ongoing email marketing campaign being launched. For more detail on exactly what this campaign involved click here, but in summary we are just sending one HTML email with 6-7 hotels in Fiji immediately after the person has booked their flights online. This story is nothing new for regular readers of this blog, so today I want to focus on digging a little deeeper into the numbers.
The full presentation I gave was mentioned yesterday, but from the slide pictured, you can see the percentage of people who purchased a hotel from the total of all emails sent. A 3% success rate is probably better than most email marketing campaigns, especially as success in this case is always a person spending money, often amounts in the hundreds of euros. As I mentioned earlier, click through rates are so uninteresting to me at this stage that I won’t even discuss them except to say that I see their value in optimizing an existing campaign, but definintely not to measure initial success of a campaign for a product like this.
But far more interesting than the overall success rate is the success rate when it is broken down into the three different points of sale. We only target half the travellers (those flying into Fiji), and we have identical offers and layout for the HTML email, but the currency depends on where the passenger is departing from. US passengers get the offer in USD, Australians in AUD and New Zelanders in NZD. Take a close look at the numbers and you can see over 4% of Australia and USA departing passengers are buying a hotel from the website after purchasing their airfares from airpacific.com and receiving by email the hotels cross sell offer.
The thing dragging down the total conversion rate are the point of sale New Zealand bookings. Only 0.6% of those emails convert. So despite New Zealand accounting for approximately one third of inbound passengers, they only account for around 5-6% of the hotels sold through the website. The first reaction is to think that using the same template and layout must be missing some pretty extreme cultural differences, but as much as we might like to say New Zealanders are different, the differences really are quite minor.
The figures on New Zealanders definitely gave me some good material for a couple of jokes when I was presenting that slide a few weeks ago, but the truth almost cetainly lies in this fact:
With such a high percentage of travellers almost certainly visiting family, the chance of them needing a hotel room is obviously much smaller. Maybe there is an opportunity to promote a different type of accommodation to these people – rather than island resorts it could be that accommodation closer to the main population centre (eg. Nadi) would be more appropriate. Or if most of them are staying with family (an example of where a market research study could yield valuable results), maybe that email could be used to try and sell something other than hotels, but which still earns the airline a decent commission.
The lesson here is that email marketing is no different to any other type of marketing in that customer segmentation is crucial in order to maximize results. There is always a cost beenfit analysis to be done regarding how much segmentation will still yield profitable results, and before changing the approach to NZ point of sale I’d really suggest getting a better understanding of these passengers and their spending patterns whilst in Fiji. The next step would then be for the airline to decide what will generate the most ancillary revenue from these passengers.
At this point in time, the answer to that last point is far from clear, but it certainly doesn’t seem to be the current hotels being promoted.