Does segmentation work? Stats and case studies
It's something to show your boss or bank manager when you want to invest in a better ESP (or more advanced software) and make segmentation part of your email marketing program:
1. MailChimp compared the average results from segmented campaigns with those from non-segmented campaigns. Segmented campaigns scored better on nearly all measures of success.
2. MarketingSherpa charted a similar story in their 2009 Benchmark Guide (link goes to latest edition), with segmented campaigns producing at least 30% more opens and 50% higher CTR than undifferentiated messages.
(Note: Both the Sherpa and MailChimp studies compare aggregated results from campaigns using segmentation to aggregated results from other campaigns not using segmentation. So it's not a direct comparison, but the trend is still clear.)
Just a couple of examples that crossed my inbox recently...
1. Fabric.com started using data collected via preference centers and web analytics to segment their campaigns and build trigger emails. Among the reported results:
"...the automated lifecycle emails helped Fabric.com double its open and click-through rates and led to a 40% increase in conversions"
2. ValuePetSupplies.com use simple segmentation to target customers with specific product offers and address new customers with a special welcome program.
They report huge ROI lifts, with ROIs of up to 12,000% (though that only includes ESP charges in the cost side of the equation I think).
3. Sears introduced a personalization program that also let subscribers choose the kinds of emails they wanted to get. The reported results of the move away from batch and blast (based on A/B tests):
"41% increase in revenue per clickthrough, a 178% increase in conversion, a 59% increase in average order volume and a 215% increase in clickthroughs."
4. The Eastwood Company introduced RFM segmentation:
"We've experienced a 20% increase in e-mail marketing revenue, a 14% reduction in the number of opt-outs and decreased e-mail marketing costs by 25%"
5. Intermix saw clickthrough rates improve from 20% to 50% after segmenting recipients by their interest in different products and offers.
6. Motorcycle Superstore classified their email addresses into six customer types and sent each different content. The result? Open rates doubled and clickthroughs tripled.
7. Shoeline.com began sending content based on RFM segmentation criteria and previous purchases. They noted that the ratio of orders to opens was 31% higher than for emails segmented more traditionally.
You split your email list into groups, where each group has some shared characteristic(s). Then you tailor your email program to each group, with email content, timing and frequency reflecting those shared characteristics.
Inevitably, this boosts responses. And the above numbers confirm this.
Anything that lifts the relevancy of your message lifts responses and protects against email fatigue.
And as the holiday/recession-driven rush to send promotional email continues, email fatigue might just be the email marketer's number one enemy.
But...you rarely see data on the net value of segmentation after you account for any additional costs incurred. Advanced segmentation in particular might mean investment in new database and email technologies.
So don't forget to look at all the numbers: balance improved responses against the costs of getting that improvement. If you're lucky, you may find that - as with the Eastwood example - total costs even fall because you end up sending less email.
Not that segmentation has to cost much (if anything), as this post and Jeanne Jennings series on Really Simple Segmentation demonstrates.
Anyone else got numbers to back up the argument in favor of segmentation?
Tags: email marketing, email segmentation
Very timely post Mark, email marketers should absolutely be taking advantage of segmentation to deliver more relevant content to their subscribers and lowering the signal to noise ratio.
Yaro Starak posted a great case study yesterday on his blog outlining how he recently ran a promotion on one of his subscriber lists utilizing segmentation tools built into AWeber.
By Tom Kulzer, on 28 October, 2008
Thanks Tom. I nearly included Yaro's post in my original list, but left it out as it had numbers, but no "before and after" comparison. A good reference to have though: thanks for bringing it to our attention.
By , on 28 October, 2008
A great post there, and is the question we all get. I can honestly say that taking advantage of segmentation is a worth while and important factor of email marketing, you can reduce bounce rates, increase click thru's and most importantly it helps to give the reader something relevant!
By Adam, on 29 October, 2008
Great post, Mark. I think segmenting is absolutely essential.
I remember when I first joined a list of a clothing company that I order from. They sent completely irrelevant emails almost every time so I unsubscribed.
A few months ago, I tried again. This time the company is obviously targeting more relevant emails to me and, presumably, to other customers as well.
They are using segmenting and getting more of my business and I won't unsubscribe this time.
By Neil Anuskiewicz, Business Development Director, on 01 November, 2008
Thanks Adam, Neil,
I hope the post allows the undecided to cast a vote for segmentation. A few concrete numbers often helps.
By , on 03 November, 2008
Thanks for this great post and the links.
One aspect seems to be missing in the discussion so far: How far do topics spread out in the non-segmented versus segmented mails? We had very different topics, very different functions in the same target group (engineers). Sometimes a nice mix adds to the value, sometimes it stops people from reading.
I'd always go with -stick to what is working and take a small sample to try something new.
By andreas.wpv, on 10 February, 2009
andreas.wpv: a very good point. I grapple with that problem myself.
I know some of my newsletter subscribers are mainly interested in email design articles. But that's not all: they also click (just less than the average reader) on other stuff.
Should they get segmented out and sent just design articles? Is it worth segmenting them out when they also value the topic mix? Can I segment them out and send the same newsletter everyone gets, just with the design articles up top in the most prominent position?
I'd do the latter if I had the time and tools.
By , on 10 February, 2009
Comments closed for this post