In my experience, the day of the week and hour of the day at which marketing emails are sent is often based on little more than the gut feeling of the email marketer and the performance of previous emails, rather than real data.
As someone who could put the anal in analytics, I think that's a rather inexact science. Surely there's a more accurate way to figure out whether the assumption is really true?
There are clearly some days of the week and hours of the day that result in higher conversion rates than others. So theoretically, if you can get your email marketing to your customers' inboxes at the time they're most likely to convert, or just before, your efforts should result in better conversion rates and more revenue.
Fortunately, there is a better way to determine the optimum mailing time, rather than using gut instinct - you can do it via Google Analytics.
However, it's actually surprisingly challenging to pull this data out of GA.
In fact, it's currently not at all trivial, without resorting to some advanced segmentation and some ingenious interface hacks that allow you to use Analytics in a way in which Google never intended.
On the plus side, no coding is required. It can all be done via your browser.
Step 1: Create your segments
In order to find the time of the day at which your site's conversion rate is at its highest you'll need to create some advanced segments to separate transactions into time period segments.
You can make as many of these as you want, but for a reasonably accurate picture three or four ought to suffice. I went for morning (7am-12am), afternoon (12am-5pm) and out of hours (5pm-7am).
If you notice that the out of hours segment seems to provide particularly good conversion rates you may want to add an additional segment for the evening, splitting the day into quarters, giving you slightly more precision.
To create these advanced segments click Advanced Segments > All Visits > Create a new advanced segment. Then click Dimensions > Visitors > Hour of the day and drag the bar to the "dimension or metric" placeholder at the top. Click the "and" link and drag a second Hour of the day bar to the "dimension or metric" placeholder.
Enter the time periods into each field to segment the traffic up according to the time period. Or, to save yourself the time and effort, just click the links below and the segments will automatically be added to your Google Analytics account.
Step 2: Finding the best time
Go to Advanced segments > All visits > Custom segments, then check each of the segments you just added or created, then click Apply.
Now click the Ecommerce button on the left hand navigation. At the top you should see the All visits conversion rate, and the three conversion rates for the time segments you just added. If one of them is better than the others, then that is when you should send your email.
Use the date widget to check different weeks, months or extend the length to cover a year or more, if you've got sufficient data, so you can double-check that the data you're observing is consistent.
Once you've identified your peak conversion rate time window, it's worth drilling-down a bit further, so go back to the advanced segments tool and create some extra segments for the time periods that fall within.
This will allow you to determine whether the highest conversion rates occur at 7-9am, 9-10am or 10am-12am. You'll want to time your mailing so that it covers all of the peak conversion rate times.
Step 3: Hack your browser
Given that we've just made advanced segments to visualise time periods with relative ease, you'd think it would be straightforward to do the same for days of the week. But you'd be wrong.
Finding the best day of the week is the particularly tricky bit! I scratched my head for quite a while on this one, and I don't think it's even currently possible to do this in GA at the moment - at least not without some browser witchcraft.
The current interface of Google Analytics doesn't allow you to compare data for specific days of the week to see which one provides the best historic conversion rate, because there's no dimension for it.
Weirdly, the day of the week dimension functionality does appear to exist within the GA software, but it appears not to be enabled.
However, it is possible to hack the GA interface to get at the dimensions Google doesn't provide in the current version, which allows you to analyse data in ways the average GA user wouldn't be able to do.
Simply install Greasemonkey, restart your browser then click this Google Analytics Report Enhancer link and Greasemonkey will install a browser hack. When you next visit the Google Analytics site you should see an additional logo next to the GA one at the top left.
Step 4: Finding the best day of the weekTo find the day of the week that has the highest conversion rate, go to Traffic Sources > Search Engines, then click the Sources button in the first column of the table. This will open a mega menu style drop down and you should see a link called "Day of the week" under "Custom variable keys". Click it, then click the Ecommerce tab. You should now see a list of days of the week, along with the metrics for each one - including the all importance ecommerce conversion rate. Pick the one with the highest conversion rate, or the one which spans a few days with high conversion rates. Combine this with the hour of the day data from the earlier steps and you've now pinpointed the theoretical optimum time to send your email marketing.
- Create advanced segments for distinct time periods and drill-down where required in order to get greater precision.
- Use Google Analytics Report Enhancer to add extra functionality to Google Analytics.
- Determine the best time of day and day of the week based on the best e-commerce conversion rate for your site. Chances are, it will differ from site to site.
- Check your data before acting upon it. Look at multiple months, just in case it's a seasonal effect.
- Use GARE to check the best days of the month. Does conversion rate go up around pay day?
- Do a split test on your email database to see if there's any improvement in its performance. If it works, and you can repeat the experiment successfully, go with it.
- If you spot a pattern, also consider trying a similar thing with PPC ads. Bid up when conversion rates are high. Reduce bids when they're low.