A comment came up on a post I read on the A4U forum which got me thinking about this.  It is an age old argument and opinions have moved on a lot over the past few years.  In the old days (sometimes I like to refer to them as the good days!) client’s were not as clued up as they are today on search engine marketing and generally affiliates worked in this space.  However, with the emergence of agencies, more skilled personnel etc most big clients will manage their own PPC strategy, whether in-house or through an agency and many will not allow affiliates to work in this space.

Is this right or wrong?

Well, there is no right or wrong answer as usual. I’m very much of the opinion that it really is down to the keyword, some keywords will naturally be cheaper for the client to convert themselves without the assistance of affiliates.  However, some keywords may well be better converted by affiliates.

For example (and the point that was raised in the post I read), who would convert a voucher term better?  The brand itself or a voucher code site?  Thinking about the intent here, the customer is specifically looking for a voucher if they type in something like “brandname voucher”.  Now, sending them directly to the client site here may actually not be the best thing to do, even though the customer is directly looking for the brand.  They want a voucher, and it is quite likely that they will want to research to find the best deal they can get.  On this basis, it would surely make sense to let a voucher affiliate work in this space and convert the customer for you.  Yes, there is a risk that they could go elsewhere but if you have a good proposition and offer then perhaps the affiliate is better placed to convert this to a sale.

Food for thought on a Sunday afternoon …..

No, is likely the answer.  However, it has been apparent over the last few months whilst taking up my new challenge at Sky that many people do not understand the process behind de-duping and why it is carried out. I think there is a huge gap in understanding and communication between clients / networks / agencies and affiliates about the process and what happens.  And to be honest I’m not surprised.  It is not because it is complex, in fact it is a very simple process but the difficulties in transparency stem from the amount of data that it involves, especially when you start to look at post impression, we are talking millions and millions of touch points.

Why is de-duping necessary?

It is necessary in order for a client to attribute sales on a last click basis to one channel / affiliate.  Because channels work on a different model, display on CPM (usually), search on CPC (usually) and affiliate on sale (again, usually) cross channel attribution is not for remuneration purposes directly, it is merely a function to understand your cost per sale across these channels.  However, for affiliates it is used directly to pay commissions and is obviously vital to do when you have a large affiliate campaign across more than one network.

Is the last click model a fair view?

This depends on how you look at it.  If you are looking at the last click in isolation then no, there will likely have been a number of other touch points prior to the last click that attributed to the sale.  However, at present it is the fairest and simplest model to attribute a sale to a channel.  At the A4U Expo I attended the eBay presentation with interest regarding their click value attribution model.  In a nutshell they look at the time between the purchase and the click to define what value that click added to the sale.  I think this is really interesting but is it right to necessarily say that a first click has less value to a sale than a last click?  Would that sale have ever been made if the customer never saw the display ad that was the first click for example?

I do not believe there is a blanket answer for attributing sales.  I think it depends on the decision making process the customer goes through, the product itself, the online marketing mix the client uses and the investment in each channel.  There is a lot of talk about multi-channel attribution and I expect it to take place in 2010 but I don’t think there will be one formula, I think there will be a number of different models, which personally I think is the way to go.

What does this mean for affiliates?

I’ve always been concerned that as more investment is put into online marketing, on a last click model certain types of affiliate will lose out.  These affiliates will tend to be content affiliates, who are adding value at the decision making stage.  I firmly believe these affiliates are fundamental in driving online sales and in order to protect these affiliates a multi-attribution model needs to be looked at across channels, and within certain types of affiliates.  I think potentially “last click” affiliates such as cashback and voucher sites need to be treated as another channel within the mix.

Lot’s to think about in 2010, which I believe will be a pivotal year in affiliate marketing.  This however is a discussion that has been had across the affiliate industry for several years now, so why has no-one done it?  It’s all about money, budgets and proving incremental value.  Ever tried telling an accountant that you need to start paying more for sales that you are currently getting?  Proving that it adds incremental volume is key to success, and to get  that the whole affiliate industry needs to back it to work.

One of the hot topics at last week’s A4U Expo was the last click model and since the Expo I have been discussing this with others, reading blogs etc. so thought I would add to the debate as it is something that I think is open to wide debate.

So, let’s have a think about how to attack this one. I think we need to look at this in two sections. Last click model within the online mix and last click model within affiliate.

Last Click within the Online Mix
Most clients will run a variety of activity through online, including but not limited to affiliate, PPC, display, email and social media. Generally the sale is attributed to the last click. So at the end of each month when I as a client am looking at what channels delivered what sales, I base this on last click.

Does this reflect the true value of each channel? Well no, but as long as we understand the impact of each channel then this is fine. Display media for example tends to be the first click in most scenarios as it is used to introduce the customer to the brand / product / offer and customers will then go to find out more through other mediums. Display generally in this kind of scenario would come out at a high CPA because less sales are allocated on last click but the reality is that some of the sale attribution is within display so the “real” CPA would be lower.

So, we should be considering the influence each channel has on a sale in order to see its true value within the CPA. The payment models in these channels are very different so I don’t believe you would consider changing this but you should understand the value of each channel within the mix to understand how to allocate spend.

Last Click within the Affiliate Channel
The affiliate channel is much more complex than most channels and in many cases there will be more than one affiliate involved in the path to sale, and perhaps most importantly, affiliates are only paid on sale which is defined through last click. So unlike display or PPC where the commission is paid on impression or click the affiliates are directly affected financially by the last click model. There has been much debate around the last click model and at this year’s Expo highlighted that.

So, should we be looking past the last click model? I think yes, however I don’t think everyone should. If we think about the path to sale it differs vastly between sectors, clients and products as depends on the customer’s decision making process. A short decision making process will have a lot less touch points than a more considered purchase. For example, I’m buying some flowers, this product is straight forward and there is not much differentiation in the market so I simply look for the cheapest or those that will deliver on time. I wouldn’t have many touch points, potentially just one. However, if I was buying broadband this is a much more considered purchase and I’m likely to look at a number of sites before I get to my purchase. I think it is mis-leading to take one client’s stats to be indicative of any other client as they will be very different.

I am yet to see a proposal where I think the last click model will work. The problem is that no client is going to pay a higher CPA so the only answer is to carve up the CPA and this leads to a lack of visibility over what an affiliate will earn from day to day, there are additional difficulties across dual network and incentive sites will not be able to determine how much they can pay their customers, therefore negatively affecting their model. The more I think about this the more I think it will not work, however, am open to ideas.

I’m starting to think that perhaps the current issue is around visibility. If affiliates had data to show where they were in the path to sale would this enable them to understand better how their site interacts with the customer and what part of the decision making process they fit into? And, if affiliates did know this could they adapt their strategy to become the last click? Perhaps the emphasis should be on affiliates to solve this issue rather than the Client?

Is it the end of the world for affiliate marketing as we know it?! I hope not, otherwise this huge Expo would be a bit of a waste of time, at least we are not all on a boat this year, last year the whole affiliate industry could have been wiped out with one leaky hole!

I’m very excited to be on this panel which Duncan Popham is leading and am joined by people who have a real passion for affiliate marketing, Julia Stent (Vodafone), Seth Richardson (DC Storm) and Robert Glasgow (Webgains).

So what is it all about? Well, the idea is that we want to form a panel discussion about the state of the affiliate industry and how we adapt to get where we want to be, which for most people I would assume is a bigger slice of the online marketing pie. We will touch on some current challenges which have come from the evolving channel such as the last click model, de-duping and transparency. All subjects that various stakeholders have differing views on so should make for a lively discussion.

I’m talking about transparency as this is something I have been focused on over the last year. I have some real concerns about transparency, not just from a client point of view but also in terms of what affiliates receive from clients and networks. This should be a really interesting and lively debate and I hope lots of passionate people will come along and pose some challenging questions for us as that’s what makes a good debate.

If you are not attending the Expo, can’t make the seminar or perhaps are not keen on voicing your question in an open forum please feel free to send to me pre-Wednesday and I will submit to the panel.

Affiliate Apocalypse Panel
Day Two
14.30-15.30

This will be my third year at the A4U Expo so starting to feel a bit like the Excel centre is a second home as have spent many hours of my life there now.  I am really looking forward to this year’s event mainly as I come from a different angle now working client side.  Previously I have been at a network so working on the stand, chatting to people, arranging meetings so often missed a lot of the conference session previously so this year I can really make the most of some of the great seminars on the agenda.  Here is what I will definitely be making an effort to see:

The Pre-Event Party

I’m not entirely sure I can make this one as the Excel centre is as far away from the BSkyB offices as you can get but I will try very hard as in previous years this has been arguably the best night.  Most people arrive on the night before and everyone is relaxed so a great time to get chatting to new people and old friends before the intense atmosphere of the conference the next day.  For those that have been there the last couple of years they will remember the strange lights overlooking the docks in the water, you have to see it to believe it.

Tuesday 13th – Day One of the Conference

9.30 am: Affiliate Marketing within the Media Mix

If I manage to get to Excel in time for this I will be heading to see Ewan talk. This is a subject I am quite keen on and is part of my role at Sky so might see if I can pose a few questions.  I have heard Ewan speak a few times now and find him engaging and with a great understanding of affiliate marketing with a rare view of how it sits within the media mix so will be an interesting talk.

10.30am: Are Networks Dying

I will definitely be attending this talk as is something I have had numerous conversations with people about over the last 6 months.  I feel that networks are under real threat and must adapt in order to add value to clients.  I have been lucky enough to see both sides so will be really keen to join in this debate and hear the opinions of the different stakeholders.

2.30pm: The Truth will Out: Disproving Misconceptions in Affiliate Marketing OR Third Party Applications: The Future of Affiliate Marketing Industry

I’m really torn on this one, I know that Kevin Edwards has some really interesting data for his talk “the truth will out” but am really interested to hear expert opinions on third party applications and their place in affiliate marketing as I see this as a key area of growth for the industry.  Will flip a coin, unless anyone wants to take notes for me?

5.15pm You Want to Tweet What?

Very interested to hear what the expert panel have to say about Twitter that we have not already heard!  Am hoping it will be an interesting talk which will spark some new ideas.

Breakout Party – needs no introduction, the party of the conference, everyone who is anyone will be there.

Day 2 – The Hangover Day (not for me of course)

9.30am: Only 4.6% of marketers can spot a good landing page, can you?

This sounds really interesting and a bit different from some of the other subjects on the agenda and something that I would like to see some good case studies on as landing pages are often the simplest way to increase conversion

12.15pm: Beat your Merchant at their Own Game

The synopsis of this panel is as follows, I think I shall hide at the back, I hope this is more tongue in cheek however.

“You’ll not only increase your affiliate revenue, but even be able to profitably outbid your merchants and competitors on AdWords and push your earnings even higher. (Just make sure that they’re not in this session too.)”

1.30pm: What can the IAB Council do for You

As a member of the IAB Affiliate Marketing Council I will be giving my support to this panel and I hope that many people go as I think it really important not only to hear what we have been doing but for people to pose questions, concerns and ideas.

2.30pm: Affiliate Apocalypse Panel

I can’t get out of this one as I’m on the panel!  I think this will be a really lively debate as we touch on some controversial subjects and issues that people are very passionate about such as transparency, last click model and the challenges of the affiliate channel in 2010.  There are some really seasoned speakers on this panel including myself of course, Duncan Popham who has been around forever I believe and Seth will offer a really interesting view from a tracking perspective.

4pm: Case Study eBay’s Partner Network

I’m really interested in the quality click pricing that eBay have introduced so will be heading this to learn how it was implemented, why and successes so far.

After this session I will be calling time on hopefully a very enjoyable and informative Expo and heading for some relaxing drinks at the Fox which has become a regular haunt for the affiliate expo.  Last year I met Matthew Le Tissier, wonder who I will meet this year…

Hope to see you there next week

The most common question I get asked tends to be around cash back sites, do they add incremental volume or not?  Unfortunately, there is not a right or wrong answer as it completely depends on who you are.  Amazon for example do not work with cash back / loyalty sites (with the exception of Nectar).  They state that the reason for this is that they feel that Amazon is a strong enough brand that customers will shop there regardless of reward and customers are loyal to them.  I think this is logical.  Just taking myself for example, I am an avid cash back user, however I purchase regularly from Amazon as I know they are price competitive, I will receive the goods within a few days and it is a simple and easy process.  I am an advocate of the Amazon brand and to date I have not been persuaded away by better cash back offers with competitors.

Amazon however are an exception in the online world, most brands do not have this loyalty or presence in the market.  This week I purchased some flowers, this product has little differentiation in the market, prices do not differ greatly, most online merchants use local flower sellers so between brands there isn’t really much to differentiate.  I chose to purchase my flowers from Interflora because they were offering 10% cash back and free delivery which looked like a reasonably good deal.  Brand did play a part to some extent as I have purchased from Interflora before and trust that they would deliver good quality flowers to my mum on the day of her birthday.  However, I would have the same view of other brands such as Tesco, M&S etc. So the cash back site in question played a key part in my decision to buy from Interflora, without that option I would have purchased elsewhere so this would be considered an incremental sale.

There is no blanket answer to whether cash back sites deliver incremental value or not, it depends on your brand and the competitiveness of the market.  If you have a product that is exclusive and can not be purchased in many other places then you should consider whether the customer would have bought from you regardless of reward but if you have many competitors then I believe the incremental value is significant.

I have worked at two affiliate networks over the last 7.5 years so I not only understand the pressures involved from this side but more importantly I have seen how this has rapidly changed.  Back in 2002 there were a few key networks, UK Affiliates (now dgm), Commission Junction, Online Media Group, Affiliate Window, Tradedoubler and Buy.at.  So much the same as it is today except for a few additions from outside the UK.

When I started in affiliate marketing back in 2002 I started as an account manager, dedicated to one rather large account which later became two accounts, but never more than two.  Because my time was solely spent on these accounts I knew their business model inside out, I knew all of the affiliates, I had history with the accounts so I knew which levers to pull to drive sales.  It was an incredibly successful partnership.  Over the last few years I sadly think that this dedication to account management has been lost from many affiliate networks, visible proof of this has been the entrance of agencies into this space and some clients moving their campaigns in-house.

So why has this been lost?  I think there are a few reasons for this but the most significant is that overrides have decreased.  It is becoming harder for networks to justify the time they can allocate to each account and these conversations happen regularly, the sign of a saturated market perhaps?  I also think that in a quest to gain business affiliate networks have shot themselves in the foot by selling themselves on incredibly low overrides.

So, where does this leave affiliate networks? It’s a tough environment out there at the moment for networks, the barriers have dropped and online agencies have got a foothold in the market, selling themselves on consultancy.  Off the shelf tracking technologies are easily available and some networks even lease their own technologies.  The larger affiliates have more buying power and can easily move direct and clients themselves are seeing advantages of taking their entire campaign or the top affiliates in-house.  So perhaps a hopeless case you might say?  However, ever the optimist I think that there is still a place for networks but they need to position themselves carefully, should they offer service and added value or should they simply be a technology provider?

I will watch this space with extreme interest but I would expect that in the next year things will only get tougher and will be interesting to see where the networks strategies take them next.

If you are interested in this topic get yourself to Day One of the A4U Expo as this is a topic being discussed and I would expect it to be a lively, heated and emotional debate.