For Blog's Sake - Don't You Know I Add Value?

Editor's Notes

• #2 For Blog’s sake – don’t you know I add value?
• #3 First time speaker: 4yrs experience as the network’s publisher representative: champion the rights of all publishers regardless of promotional methods/ size/ revenue or reputationIf a man can be known as nothing else, then he may be known by his companions
• #4 First time speakerPertinent to introduce myself | establish credentialsAnswer the following: Why did I ask to speak
• #5 The topic is understanding how influential bloggers are
• #6 Common theories:frequent advertiser requests: find me long tail publishers, expecting sales/ sales funnels
• #8 Let me introduce you to your average affiliate marketer; not because they pillage your treasure or because they have hooks for hands but because I believe they only see half the picture.There was a recent article on the A4u about whether affiliate marketing is too measurable, if anything the opposite could be argued. There are too many things in affiliate marketing that can’t be measured, most importantly would be “influence.”Influence is defined as “The capacity to have an effect on the character, development, or behaviour of someone or something”
• #9 It’s challenging: bloggers are hard to engage because the value they deliver is frequently misunderstood and it’sdifficult to quantifyMillions are spent on tv/radio/print/display but retailers argue about 1 or 2% margins in the performance channel or decide to only pay on new customersEveryone wants a bigger longtail/broader range of partners but no one knows how to justify the costs to their bosses
• #10 I want to help you justify to your bosses, your colleagues, your board members that affiliate marketing has more to it than just converting salesI want to provide you with data that actually shows the influence that long tail publishers have and why they’re worth more investment than 5% of a transaction value
• #11 My goal today
• #12 This goal isn’t an easy thing to prove but it would be impossible without providing a bit of clarity on what I mean by “premium blogger” I’ve got some examples on the screen behind me but a couple of requirements will also help to cement this subjective term into something more measurable.Premium bloggers have:Alexa Traffic Rank (UK) < 100,000Style and rich contentAn established brand identity
• #13 Basic blogs haveNo alexa rankingA subdomain of their webhostsA basic templated layoutLinks to other non-related blogs
• #14 Donald Rumsfeld once saidThere are known knowns. These are things we know that we know. (for example we’re all at the expo)There are known unknowns. That is to say, there are things that we know we don't know. (for example how many people are in the expo at this exact time)But there are also unknown unknowns. There are things we don't know we don't know. (well I can’t give you an example of that because if I did, then it would become a known unknown!)Still with me?I’m going to begin with some known knowns and then work through some known unknowns so that they become known knowns and we can all leave this session knowing more than when we came in!
• #15 Here is a known known, it’s the typical sales funnel that a customer supposedly goes through when making a purchase.Alongside this funnel is what we, as marketers are taught to believe is the online channel best suited to engaging a customer in their different stages of a purchase.There are varying versions of this with minor adjustments but essentially display is considered the best promotional type for making a consumer aware of your product/service/brand and helping them form an opinion of what you stand for/ how good your products are.We then have search and display demonstrating an equal ability to aid potential customer’s consideration of a particular product or service. Taking an example of a travel purchase, this might be exploring the different countries/cities that they can take a city break to.Once options are considered, a customer decides on a preferred product or service and completes a purchase. This is where search and affiliate marketing are considered to be most effective. Affiliate marketing is suggested therefore as the “conversion” channel.
• #16 If we focus purely on the affiliate channel for a few moments, you can see a piechart which shows the number of times affiliates were involved in a sale, including those where they didn’t secure the last click (with a DIY retailer). The notion that affiliates are purely conversion specialists therefore doesn’t stand up to huge amounts of scrutiny.
• #17 With my slightly biased affiliate hat on, I might suggest that the sales funnel be re-written as follows. Awareness and opinion are influenced by television, tube advertising, mainstream press and you will find affiliates in all of these places!Truth be told, I think every channel has a part to play in each of these stages but I still think there’s a missing part of the jigsaw and here’s why...
• #18 This is a situation that a lot of people will sympathise with, I moved into a new place and one of the door handles wasn’t exactly in tip top position. Now I have to confess that I’m completely useless when it comes to anything DIY related. Normally, I get someone else to do it but as I’d only just moved into a house share, I didn’t want to make a bad impression so I decided I would try to mend it myself.
• #19 In true consumer style, I started off my search on Google and stuck to the basics; buying replacement door handleI click on the first link (naturally)
• #20 This is the page I arrive on. Suddenly I feel overwhelmed as I realise I know absolutely nothing about doorknobs (not door handles as I originally wrote) and that I probably need to work out which one is right for me
• #21 Back to google and I’m now trying to understand the intricacies of the door knob world.
• #22 I turn to those in the know (ehow) and James from J&C home repair provided me with 70 seconds of expert knowledge. He taught me two things:All door knobs are the samePurchases should be made on personal preference
• #23 Feeling secure that I can’t actually get this wrong, I turn to Google images to find something that I like.Then I see this gem of a door knob, stylish, cool looking and a bit funky
• #24 Clicking through I come to learn that the product I like is “modknobs”
• #25 I’m straight back to Google and in search of a bargain. Turns out there are no ways of getting a cheap modknob and so I click through and complete my purchase. As far as any funnel analysis is concerned, I have arrived on a branded SEO term. Truth is my journey was completely different to how a click path analysis would portray it as I was actually heavily influenced by the editorial site materialicious.
• #26 When you then see pie-charts like these I end up with loads more questions about the data:Does Direct2site really get that much traffic or are there other factors at play? It all comes down (in my eyes) to a question...
• #27 The only way of really knowing is to stalk everyone whilst they browse the internet or to stick trackers in everyone’s eyes and monitor what they look at.Well actually, this has already been done; cue stat.
• #28 So with the goal of turning a known unknown (how much insight is missed by click tracking) into a known known I set up a collaboration with a premium blogger; an experiment if you will.We didn’t quite strap people in chairs and put trackers in their eyes, but we did attempt to understand how people behave on the blogger’s site and what their purchasing habits were afterwards.
• #29 The premium blogger of choice wereAigua media; they have:11 Internationally renowned blogs 2 Unique social shopping networks12 twitter sites
• #30 This is the post view model, it is normally associated with display and bannersIn essence, if a customer views a banner on a publisher site and later goes on to purchase without having previous or subsequently click on other marketing channels, then they will be rewarded with commission.We tweaked this method slightly, we implemented post view tracking on all pages for Aigua media’s portfolio of 11 websites. We monitored this for four advertisers and then ended up with visibility on how many people pass through the site and later purchased with these four advertisers.
• #31 The results are staggering.In the orangey/salmon colour, you can see all the transactions we tracked from users visiting any of Aigua’s sites vs all the transactions that Aigua actually made by securing clicks from their users. There’s two initial interpretations you can take at this early stage:Aigua media struggle to get their users clicking through to retailer sitesAigua media are hugely undervalued deliver a shed load more sales than they get credit forThe only way to really know is to mine further into the data
• #32 Just by way of comparison (because these results blew my mind) – I tested a cashback site.I put post view tracking on for 2 days and compared the number of site visitors against the number of people that clicked and purchased.Unsurprisingly, almost everyone on the cashback site did click through and purchase whilst so few of the visitors clicked through from Aigua(this is obviously because of the different business models but it’s important to sense check these things when you get results like those in the graph before)
• #33 Going back to the Aiguatranasactions; A point of note is that in around 5% of these transactions (it varies per retailer) , Aigua’s site visitors then visited a different publisher site. It’s possible that in these cases you could say that Aigua could have done more to convert these visitors into “clickers” and therefore secured the sale themselves.Given that this only happened in 4% of transactions, further weight is given to the argument that actually Aigua do add more value than is immediately obvious.It would be appropriate to mention that in no way am I saying that Aigua influenced every single one of these purchases, it could have just been a happy coincidence that people visit Aigua and then purchase later. The real interest is understand the difference between a visitor and someone who’s been influenced.
• #34 One way to consider influence might be latency (or in layman’s terms – the time between the sites being visited and the purchase actually happening)
• #35 It then comes down to where you set the bar. For example, you might consider that anyone who purchased on the same day HAS been influenced by Aigua’s sites. This would suggest that only a third of all the sales we saw in the first graph are actually relevant. Even so, this is still huge amounts more than they’re currently rewarded for.
• #36 You could set the bar higher, You could say that anyone who purchased within the same hour was influenced by Aigua and everyone else should be ignored.You’d still loads more transactions than Aigua are actually being paid for
• #37 In fact, if we consider every single transaction and break it down into weak influence, medium influence and strong influence, the disparity between what they’re paid and what they seem to have influenced is pretty incredible.I had a brief look at how this might change their monthly earnings, and it’s unsurprising to hear that this would have increased their earnings by more than 25 times.
• #38 If we put the slightly more scientifictime argument to one side and focus on the more subjective method of looking for similarities between the departure page on the affiliate site and what was purchased, we might get a different set of results.
• #39 Here are some examples I found whilst going through the data: Some might consider this to be a slight correlation between the page viewed on one of Aigua’s sites and the product the user actually purchased from Debenhams. I’ve decided to label transactions of this nature “weak influence”Weak influence, is where there is a loose connection between the product viewed and the product purchased, this could be on a category level, a colour level etc...
• #40 The next example is a weaker correlation between these two products but a strong correlation on a brand level; the Aigua site was linking directly through to Debenhams.Strangely enough, the user didn’t click but did purchase a make-up product later that day from Debenhams. I’ve defined instances like this to be “strong influence”A couple of caveats are worth mentioning here: this is all very subjective, you may have looked at these similarities and thought “NO WAY”! Those products aren’t similar at all:One’s an eye liner and one’s eye shadow, Edwyn obviously doesn’t understand fashion (which would be true)It’s also worth saying that I could only look at a smaller sample of data (around 5%) for the similarity correlations as not all retailers were able to tell me what was purchased.
• #41 Monitoring the data purely on similarities gives less dramatic results in the sense that the margins of under-valuation aren’t as strong but at the same time – it’s still saying they should be paid more than double what they are currently paid if you’re accounting for unseen influence.I’ve put in a caveat here because the page which a user leaves an affiliate site may will not give a picture of all the pages the use looked at whilst on site. If anything, it’s likely they looked at more sites and were more heavily influenced than this graph suggests.
• #42 If we consider that we see almost all of the value that cashback sites deliver, but a very small amount of the value that editorial content sites deliver; what would be the effect of applying an “influence factorial” into the average programme make up?
• #43 On the flipside, we’ve only done this experiment with one site. In the same way that were lots of sites involved in my doorknob shopping experience, it’s likely there were other sites that failed to secure clicks and consequently they are undervalued too! This is especially clear when we see how many of the initial tracked transactions Aigua did not influence.Essentially, what we’ve managed to do with this post view tracking is bring forward the point at which we’re able to capture data about a user’s journey to purchase. In an ideal world, we would put post view tracking on everyone’s sites and then build up a solid understanding of who’s visiting where before a purchase is completed
• #44 What’s reassuring is that some advertisers are already aware of the unseen influence and actually do reward affiliates accordingly. Take this quote from BT as an example
• #45 It could therefore be time for advertiser’s to follow on from BT’s example or risk premium bloggers adopting the attitude of RON RONRondoe (semi famous rapper in Kansas City) and expecting advertisers to “pay me or pay me no attention!”