Posted May 20, 2011 5:53 pm by with 20 comments

Tweet about this on TwitterShare on LinkedInShare on Google+Share on FacebookBuffer this page

We spend a lot of time around here discussing the concept of whether or not social media marketing is worth the effort. We’ve seen reports that show that website traffic from social media networks is pretty low and I’m about to show it to you again. But this time, I’ll take it a step further thanks to the folks at SeeWhy.

SeeWhy studied a sampling of 60,000 ecommerce transactions across a variety of sites in February 2011. The first thing they did was chart where the traffic came from, but not all of the traffic. Since running an online store is about selling things, they looked at only the people who actually loaded items into their shopping cart. Here’s what they got:

As you can see, email brought in the most traffic, with direct and search right under that. Social media came in at only 4.3%, not great, but better than display advertising, so that’s something to think about.

Next they took that info and charted the percentage of people from each source who completed their transaction. In other words, the people who handed over their credit card number and hit submit.

Here you can see that email and direct hits to the site still resulted in the best conversions, social media picked up 2.11% of the overall, but look at display advertising. Yikes, only .53% conversion rate? Maybe you should be spending money on your Facebook page and not on display ads.

At the end of all of this, SeeWhy does put in the usual disclaimer. Your mileage may vary and different types of ecommerce sites will have different results. While I’m sure that’s true, I imagine the shift wouldn’t be huge.

What I get from all of this is that email is still king when you’re trying to get people to buy from you. So why isn’t email marketing in the news more often? Maybe because it simply isn’t as new and trendy as social media marketing. Perhaps one day, we’ll see the flip, social media will be the old school method we take for granted while holographic email messages will show up as 2% of the pie.


  • Well, thanks for this interesting survey. What I understand is that Social Media might be a good channel to reach new potential customers, and have new subscriptions to the newsletter. And that emailing newsletter is the best in sales conversion (which can seems obvious as you are talking to your community interested in receiving news from you -> and email can offer good call-to-action)

  • This conversion data set has a flaw in that email marketing is an active marketing campaign against a list of interested parties and all other channels are not necessarily actively seeking a purchase. For example, a daily Woot email marketing campaign will naturally convert at a higher rate because recipients have signed up to receive alerts when new products are on sale.

    Direct visitors (although familiar with your brand) may not be in need at the moment. Social media conversations may not be sales related at all. SEM is advertising, usually lower in conversions than targeted list recipients. Display advertising is even lower than SEM, by nature. Links include researchers as well as buyers.

    Great post and lots to consider but I can’t help but point out the flaw.

  • Tim

    I am in the same boat as Ash, but would like to go a step further.

    Another aspect to consider is the type of offer that is being marketed within each channel. We push special promotions (extra 10% for this or a loyalty reward) geared towards the type of traffic we are targeting. Our banner visitors may get a more attractive offer because it is passive traffic in nature.

    I was surprised to see that SEM was as low as it was in this pie but also proves once again that social media is a viable channel for conversions.

  • Larry Chase

    It would be interesting to know what the data source was for these results. Are they available?

  • Cynthia,
    For my site it’s people coming from my YouTube channel/videos. It’s not even close.
    My site is kind of niche, so your milage may vary.

  • Interesting figures. Didn’t expect email to lead the pack and exceed conversion rates from direct hits.

  • Ash
    Thanks for your comments on this. I’m not sure that I agree that the data set has a ‘flaw.’ The data itself is good. What you are pointing out is that it would have been interesting to segment the data into customers vs new business, which would reveal a different aspect, and email wouldn’t feature at all.

    We know that it may take many visits to achieve a conversion, not a single visit. So new traffic may not lead directly to a sale, but if it leads to a micro-conversion, such as a newsletter sign up, registration or social login, then this is very valuable. So perhaps we should be looking at paid ads as a stepping stone on beginning to build a relationship, which ultimately will probably only turn into a sale after a sequence of touches. In the majority of cases, this data shows that the last touch will more than likely be an email.

    The lesson that I take from this data is a reminder that the majority of sales come from people you already know, and who already know you. This is intrinsically obvious, but we often forget it in our relentless quest for more traffic.

    Charles @webconversion

  • “Maybe because it simply isn’t as new and trendy as social media marketing.”

    That is an excellent point! Social media was the flashy, shiny new marketing toy. E-mail marketing was suddenly perceived as old and stale. Now that social media has matured a little bit, I think more and more companies are realizing that e-mail marketing still has a place and value in their online marketing campaigns.

  • Cynthia Boris

    Just from my own personal habits, email would be tops for conversion. Very often, I make a purchase based on an offer sent to me by email, close or equal to this is my directly visiting a site I know to buy something.

    I’ve only just begun paying visits to site and only have rarely bought anything from a social media link, but it’s happening more and more, so if I’m a typical buyer, then my numbers jive.

  • Very interesting post.
    First of all, it is not because there is a new channel that email is totaly dead for ecommerce.
    I’d like to know some little things:
    What are the sources which make email subscriptions ?
    Also, what about the same datas if we only take new visitors or first purchase?

  • Very nice diagrams and every things looks so clean cut.

    But there is so much more we’d need to know to ascertain all the kinks in the underlying figures. Naturally email and direct visitors are going to get to the shopping cart, it’s pretty much why they’re at your site in the first place – they’ve been led. Display ads are fraught with problems as so many users don’t know how to optimise a landing page and they mislead visitors by having so wide a keyword reach, social media likewise.

    Then of course: was the shopping cart easy to use, was shipping cost added at the last instant, do products get shipped within the visitors time-scale?

    Interesting post, in that it should instigate SEOs and the like to re-look at their ad campaigns and then on-site usability.

  • Search Engine Optimization Langley

    I agree with many of the posters above. Not knowing the data sources, not knowing the product type and the comparison of a existing list (existing customers) vs the aquisition of new customers is flawed at best.

    People on an email list have already made a conscious decision to listen to the retailer and respond to offers. People coming in by SEM or direct navigation (I’m assuming they’re referring to type in traffic) are generally in a research phase and haven’t made up their minds yet.

    That’s like a bricks and mortar retailer saying our existing customers buy more than our new customers that have never been here before. More of a “DUH” moment than anything else!

    • “That’s like a bricks and mortar retailer saying our existing customers buy more than our new customers that have never been here before. More of a “DUH” moment than anything else!”

      Interesting, though, that there is so much talk and focus on the ‘new customers’ (ie. social media) versus a tried and true method.

      To me, I prefer to focus the limited time I have every day on the part of my business that gets me the biggest piece of the pie – the subscriber and customer lists.

      That’s just smart business I think.

  • Your headline is misleading.

    These graphs do not show conversion rate. Conversation rate is a ratio (or percentage) of visitors who complete a desired action — in this case, reaching the shopping cart. The conversion rate for search traffic, then, would be calculated by dividing the number of search visitors who reached the shopping cart by the total number of search visitors.

    Your first pie chart shows each Source as a percentage of total traffic. An important metric, but not a conversion rate.

    Your second pie chart shows each Source as a percentage of total conversions. Again an important metric, but still not a conversion rate.

    If you want to compare the conversion rates of one Source against another, the best tool is a bar graph.

    Email may in fact have the highest conversion rate. It will vary from campaign to campaign, product to product. But one thing is for certain: These charts have nothing to do with conversion rates.

    • Thank you for pointing this out! I was going to leave the same comment. I’m sure there are hundreds of people that didn’t read the stats correctly and are applying the wrong information to their marketing plans.

      It pays to understand stats!

    • The original article from which this was taken has an additional chart, which is a bar chart and shows the conversion rate for each traffic source. Somehow it didnt make it in here. You can see it here:

      Charles @webconversion

  • I’d like to see the breakout of what traffic source got visitors to subscribe to the email list in the first place.

  • I think these statistics are extremely misleading. The company that provides them SeeWhy deals in remarketing of abandoned shopping carts. And in the definition of “traffic” they are apparently only talking about sending email to people “who have put something in a shopping card.”

    So, yeah. How would you even contact somebody like that, other than email. Adwords has a remarketing concept that is somewhat similar.

    But I think the average person would think it is trying to show that 67% of traffic comes from email, when in fact it is only saying that 67% of traffic of people who have already put something in your shopping cart is coming from email.

    Totally different from saying that people will first come to your site 67% from email.

    Very misleading.

    • Jeff

      Just to clarify here, the specifically excludes any returning visitors that might have been in email remarketing campaigns. It would indeed be misleading to include these.

      I have no axe to grind about top converting traffic sources, but like any one interested in conversion, I am a data junkie. I did some work in market research before becoming a marketer, and I guess some things don’t rub off.

      The data in this case doesn’t seek to represent which traffic sources drive the most traffic, but rather to take a different view, and look at how far down the conversion funnel visitors from different sources get. The results are in fact somewhat obvious when you think about it, but weren’t before I had done the analysis. It’s a bit like analyzing sales data to find that ice cream sales go up in the summer!

      Of course email has a high conversion rate – these are returning visitors and previous buyers. But I hope that by doing this analysis people are analyzing their own data more. And perhaps thinking about how it takes a sequence of micro-conversions leading to an eventual sale, and how each of the different channels can work together to get the conversion.

      Charles @webconversion

  • I am trying to rank this site- better. Can you suggest some good ways for it?