The source that KISSmetrics referenced was an article by Huffington Post, the grandly titled “100 Fascinating Social Media Statistics and Figures From 2012” (20 of which relate to Facebook).

Tucked in the list was the specific statistic I was hunting down, in a reference to an article by Business2Community about “Facebook Marketing Statistics You Need To Know”.

This link led me to an article by HubSpot: “12 Revealing Marketing Stats About Facebook for Business”. This is a taster selection of social media statistics from an ebook by HubSpot collating “47 Handy Facebook Stats and Charts” (the largest ever number of Facebook stats in my journey so far).

This time I had to register to download the full research, so I did, thinking that my ruthless ‘source hunt’ would be over, and I would finally get to drink directly from the font of knowledge.

Instead, however, in the ebook I found the statistic in question (“77% of B2C companies have acquired customers from Facebook”) being quoted and attributed to “The State of Inbound Marketing 2012” report, again by HubSpot. After parting with more of my precious personal data, I had finally downloaded *the* source.

So, how did this figure come about? HubSpot surveyed “972 professionals who were familiar with their business’ marketing strategy”, 72% of which were in B2B. The size and industry were widely varied, but there was no cross-tabulation to reveal which industries/size acquired more customers from Facebook (I assume this was due to the small number of participant segments that would not allow for statistically significant results).

The question was phrased as “Have you ever acquired a customer from the following social media / blog channels?” (presumingly presenting a list of social media, including Facebook).

So, is the reported figure a lie? Not really…

But is it a “fact” that we should trust? Is it, in other words, misleading? Everyone is free to chose their answer but I would need a very huge pinch of salt to make anything out of this figure. I assume there will be many marketers that read this statistic and use it as leverage to make or force a decision to create a Facebook page for the companies they work for. How would they respond if the ‘facts’ were presented like this:

“77% of the 272 B2C marketers we talked to claimed they have acquired at least one customer from Facebook”

I have nothing against Hubspot and the companies mentioned here. It is common practice in every industry to popularise findings of studies in the form of statistics. And it is also common to repost, re-tweet and regurgitate these statistics until the original knowledge generated is not relevant anymore.

Next time I’m looking for useful social media statistics I may take Dilbert’s advice and simply make the numbers up!

The change means all new campaigns in Google Adwords will be able to target different devices in different locations at different times with different bids, from the same campaign.

Google’s theory for introducing Enhanced Campaigns is sound; it is all about being able to deliver the right ad to the right person at the right time.

This would mean that a user who is on a mobile near a restaurant and is searching for places to eat will be able to see a tailored ad in her mobile search results. If they use mobile search in the evening, the ad may include a different ad creative with a map, a phone number and importantly for this more qualified lead an advertiser will be able to bid higher. Whereas someone who makes the same search in the morning on a desktop and is 30 miles from the restaurant may include different ad extensions with site links and creative focused on deals for making a reservation for the future and a lower keyword bid or position on the landscape.

This is a single example but the principle behind this is that an advertiser can change time scheduling of an ad, by location and device in a single campaign. Keyword bids on the Google Adwords landscape can be changed by these three elements and importantly an advertiser can also schedule different ad extensions such as site-links, call extensions and offers by these three parameters.

Couldn’t we do this already?

In the most case, this is true, we could.

As an advertiser’s agency we would build out separate campaigns for mobile search, have different bids for these keywords and we could target different times of day for these campaigns.

For example: if an advertiser wanted to advertise by different device and had a call extension which should be visible only when a call centre is open. Best practice was to copy the campaign and create separate versions for mobile search and tablet search, creating 3 campaigns for desktop, mobile and tablet. To have one add showing a call extension phone number when a call centre is open and an ad without this extension when the call centre is closed these campaigns would have to be duplicated again. This would entail scheduling the campaigns so a different set of ads were visible when the call centre was open/closed. It does mean the same keywords could be in 3 different campaigns (split by each device type) then these would be split into a further two campaigns to facilitate the different scheduling of call extensions on/off. This totals 6 campaigns instead of just one for the same set of keywords. If you do this for an account which has 10 campaigns there is immediately 60 different campaigns before you even start to look at targeting different locations or times of day for bidding. Therefore, this could mean substantial account management time in terms of adjusting bids and ad copy across targeting options, as well as potentially making reporting more difficult.

The changes also add some new useful features such as site-link reporting and ‘smarter’ ads which can automatically tell which device is being used, showing the preferred ads and ad extensions for these particular devices.

How does it work?

The system will now work by using bid multipliers. The bid multipliers will be set by taking your base bid and setting percentage increase/decrease by targeting options of device, location and time of day. The desktop/tablet keyword bid is the base rate bid and always on. You can bid up to 300% (or down -100%) on your mobile campaign. Location and time schedule targeting can be increased or decreased by -90 to 900% but never paused.

Is there any drawback or loss of functionality?

In short, yes! From a targeting (device, location & scheduling), reporting and bidding perspective there are some changes to functionality.

Advertisers can’t opt out.

The enhanced campaigns will automatically have advertisers targeting all devices for all keywords.

• Previously advertisers could simply un-tick devices they don’t want to be visible for.
• Advertisers could also have a specific message for a location in a separate location campaign, currently this is not possible.
• Advertisers could ad-schedule their mobile activity with a different time strategy for mobile, desktop and tablet.

Additionally Google Adwords advertisers with a focus on mobile search will no longer be able to target different mobile operating systems or device types such as android or iphone.

This adds to the concern that there is a some loss of control significantly as Google now views desktop and tablets as the same landscape so there is no bid difference or multiplier between the two (especially considering tablet bids could be up to 30% lower for the same position & different day part strategies).

Bidding: Advertisers will not be able to cap spend by device and how the different bids by device will appear in the interface will be interesting.

We are yet to see if we can get this information in reporting, in the most part it appears reporting will be grouped together, so how mobile performs at different times of the day as compared to desktop/tablet will not be possible, yet.

Reporting

Advanced reporting does have its benefits, including allowing us to see more consumer actions classed as conversions. This comes in the form of ‘goal types’, where we can easily count calls, length of calls and app downloads as conversions within Google Adwords.

Reporting some metrics such as quality score, however, will now be a “blended average” across all devices, instead of allowing us to see the differences in quality score for keywords on different devices, making it hard to track any real improvements or reductions in quality score on specific devices.

What if an application we are promoting is only available on an iPhone and not android or vice versa?

Google does claim the new system is smart enough to tell what operating system and device type users have, therefore if there is an app download link it will only show to those users using the right device type for the app. That is great although there is a loss of transparency to see if this is working correctly.

Our perspective:

Overall this should be a good thing, making campaign management easier, faster and efficient. In the words of Google, they want to be: ‘making the web work for you’ and this development should help users, advertisers and agencies in the long run by streamlining the way we micro-target potential customers with a message relevant to them. It is designed for the future of seamless digital cross-device engagement, also being able to schedule elements vital to your business such as call extensions. This is something the interface has been crying out for, for some time.

However, despite previous best practice advice, the new enhanced campaigns structure is there to accommodate easier management of cross-device landscapes and although we fully anticipate the migration to be difficult, we believe that taking advantage of these changes for our client accounts as soon as possible will prove beneficial in the long term.

Reporting flexibility and transparency will also need to improve as the system need develops as advertisers who base strategy around the performance of these elements will need to see this data.

As with all new releases, the first phase of development will uncover areas of improvement which are needed. Google have been very good in the past in listening to feedback and developing their products. This product release will be no different. There are tools already being rolled out to support transition of campaigns and new products which will be in beta testing that can be used following the development of the enhanced campaigns are already being launched.

It may take time to develop the bidding strategies most suited to each advertiser and to understand exactly how the bid multiplier will work with multiplying several different elements at the same time. Advertisers will have to ensure the different elements are considered and put processes in place for bid management.

So, this is an exciting new development which is only going to support our account management, in time.

The interesting thing is that both areas are fairly co-dependent – in fact they could be exploited as part of a fully integrated digital marketing strategy.

First up is our old friend content strategy.

Digital marketing has a habit of over-using terms until they lose all their meaning (see also ‘social media’ and ‘big data’) – and content strategy is well on this road.

It is normally understood as being analogous to the kind of editorial strategy a magazine might use – except that a fully formed digital content strategy should encompass site content, email marketing and social media. And of course, content itself could be text or video or user generated content like reviews or competition entries.

This approach is incredibly powerful – fresh, relevant content will help customers to buy your products but is also exactly the kind of content that Google, particularly post Panda and Penguin, loves to see.

For me, I think content strategy can also inform how we develop one-off campaigns – and I’m sure I’ll return to this theme later in the year. Essentially though we need to start thinking about audiences and content rather than reductively about channels and clicks.

The other area is the online customer journey.

At the heart of this is our old favourite, multi-channel attribution. Google’s research suggests that the customer journey and specifically the path to purchase is becoming more complex, taking longer and including more websites including social recommendations.

Google’s ZMOT study found that the average number of sources used for a purchase had doubled in a year, from 5.2 to 10.4 sources.  Here for reference is the breakdown by category:

‘Last click wins’ has always been an inaccurate way to measure attribution, but it is becoming more useless than ever.  It’s not only wrong but an increasingly dangerous strategy. The new game in town is about getting the first click, then using retargeting and email to bring your customer home.

The problem is that attribution modelling has not been an area of conspicuous success over the past few years. I remember Avinash Kaushik talking about some models that he had seen digital marketing agencies develop as “basically shit we made up”.

It is really helpful that Google is now rolling out the attribution modelling features of Google Analytics Premium to standard accounts. Also on the road map is a new feature where Google will create a custom attribution model for your business based on an analysis of your actual data. I’m not necessarily saying that Google’s algorithms will definitely come up with a better answer than a room of digital marketing planners (OK, to be honest I am!)

Anyway, getting a good model is only one part of the puzzle – once you have worked out how the channels interact, you can start to allocate budgets more efficiently across those channels. An important part of this will be optimising the customer journey across different devices as mobile and tablet usage continues to soar.

The final challenge will be a creative one – which brings us back to content strategy. Winning in this new landscape will require the ability to build engaging narratives that reflect the customer journey and stretch across platforms and channels. And to do this, we need to stop thinking in silos like search marketing, or social media, or email marketing, and start thinking holistically about digital marketing in the round.

Dubbed to provide an “effective and clever approach to cross-channel marketing”, Rich Ads In Search really do what it says on the tin. I first ran the original RAIS in summer of 2011. The results were great. The CTR was 2.4 times better on brand terms and sales followed.  However the billing process could have been easier.  Three years on from the initial launch, this has been sorted out – the ads are now priced on a CPC basis and no prepayment is required.

Here at Harvest, we re-visited RAIS as soon as they reappeared in the post-Search Alliance era. So I was curious to see who else is running tests. And surprisingly enough the answer is: hardly anybody.

Even Christmas-focused gadget stores and luxury brands are not interested.  Amongst the few examples we could find were a bacon-loving new-kid-on-the-block and a toy super-brand (although Lego are using the format to display their logo rather than a video).  Presumably everyone else is happy enough with their paid search performance and sees no need to experiment!

Our own data showed CTR rates up by 52% for RAIS with videos in 2012. And this is for big household names, which normally show exceptional CTR% anyway. What’s more interesting, conversion rate improved too – by 27%! Which meant the CPA came down by approx. 10%. Any ad that can do this to your brand is worth shouting about.

But here’s the surprising thing: whilst video ads in search generate a lot of clicks, hardly anybody clicks on the video itself. In fact, the majority of users still clicked on the underlined title and just fewer than 14% clicked on the deep links provided. The difference was tens vs. thousands.

So this is strange, because we all love watching online videos. But perhaps we just aren’t quite ready to interact with a video play button on a search ad?

So what is it that makes video ads in search work so well?  It looks like the CTR% doubled just because… there was a video thumbnail image next to it? It appears to be so. So maybe the famous toy brand above actually got it right – skip the video and just have a logo. I believe there will come a time when an typical user will look out for the “play” button, but we might have to wait a while.

In the meantime, Google are slowly opening up the gates to people who want to have expandable video joy on the SERP’s, but you have it be in categories like Music, or Games, or Movies to get it. Will they ever open it up to the masses –time will tell.

Until then, keep your eyes open for video ads in search – and feel free to buck the trends by clicking those “play” buttons!

You’re posting an entertaining mix of links back to your site, photos and the occasional piece of video.

So the question is, what kind of interaction levels should you expect to be getting?

To get to an answer, I’m mashing together a couple of recent pieces of research.  The first, from an article by Robin Grant in Techcrunch looks at EdgeRank – which is the algorithm Facebook uses to work out which updates make it into your news feed. It’s pretty clear that distribution of content by brands has dropped by some 40% in just three months – and that the drop is fairly uniform across even the most popular Facebook pages.

What this means is that if you have 100,000 fans of your page, on average Facebook will only show each of your content updates to 10% of them, or 10,000 people. And – not wanting to spoil anyone’s Christmas – that graph doesn’t look like it is levelling off any time soon.

So that gives us a number for the reach of each page.  The other number we need is the average engagement level – and for that we can turn to the excellent edgerankchecker.com, who have a recent post on engagement levels by type of content.

Here’s the relevant chart from that post, showing changes in engagement over September and October 2012.

This by the way chimes with what we’ve been hearing for quite a while – that photos work much much better on Facebook than video. And that might be down to the number of people who access Facebook via mobile – the last thing you want to do is spend £3 of your own cash to watch some hilarious clip on YouTube.

Piecing these two pieces of data together, we can get to a quite crude benchmark for how you’d expect different kinds of content to engage with your fans.

So does this work?  OK, here is a highly unscientific test based on a sample of, er, one.

I have chosen this recent post from Ocado simply because Ocado has almost exactly 100,000 Facebook fans, which makes the maths pretty straightforward.

It’s a photo (obviously) so I would predict 60 engagements – and in actual fact it scores 65 likes, 12 comments and 3 shares giving a total engagement score of 80.  So on this basis, you could say that this post has performed slightly ahead of expectations – which I guess you would expect given that it’s a nice topical posting.

Of course, this only applies to organic postings on Facebook – you can always increase the reach of any posting by using Facebook Sponsored Stories and this will clearly boost your engagement scores. And – like any benchmark – it is only an average. If you have exceptional content you can blow these figures completely out of the water – as President Obama proved with this, the most liked photograph in Facebook history.

A great introduction to this subject is a recent webcast from Basile Fattal at Philips  which gives a huge amount of test data – this is well worth half an hour of your time.

I was working as a copywriter for an agency that held the Labour Party account and it was my job to write all the long direct mail fundraising appeals on behalf of Mr Kinnock.

Our strategy was to build up the supporter base during the run up to the election and then send an unprecedented four mailings during the three weeks or so of run up to the election itself.

And actually, if I say it myself, the fundraising bit was fairly successful… the other bits of the campaign not so much.

We were at the time completely in awe of the fundraising techniques being developed in the United States. On the other side of the Atlantic the fundraising period at that time lasted for some fifteen months before an election (now it barely ever stops.) Consequently in the US they had enough time to test and refine their processes.

I digress slightly, but I remember at the time reading a book about political fundraising which explained how the Republicans had tested a vast number of bought in mailing lists trying to find one which best betrayed political allegiance. After extensive testing they discovered that absolutely the best list for the Republicans was the membership list of a rather strange organisation – the Shammy Leather Owners Club of America.

So there you have it – the activity which best betrays right-wing tendencies is the simple desire to polish your car on a Sunday morning.

Anyway, I haven’t been able to shake off my fascination with politics since then. Like many people in marketing, I’ve come to suspect that elections are not about issues – instead they are won or lost by good or bad marketing.

So in spite of my good efforts, nothing could save Kinnock from his appalling Sheffield Rally (“We’re all right! We’re all right!”) Poor event marketing shepherded in the grey years of John Major.

That was the same election where the Tories rolled out a classic poster – “Labour’s Double Whammy”. Not as powerful of course as “Labour Isn’t Working” from the second Thatcher victory.

Obama’s victory four years ago was generally seen as down to his campaign’s mastery of social media. However the element that impressed me was his use of tactical SMS messages to get the vote out on election day. It turned out that the Obama campaign had captured millions of mobile phone numbers – partly by offering free entry to events like Springsteen concerts in exchange for mobile numbers.

Just as the 2012 Olympics belonged to Twitter, so I think that the micro-blogging site will be hugely important to campaigning in this US election. Twitter’s reach (around 16% in the United States) is puny compared with Facebook, but it is the place to set the agenda amongst opinion formers and journalists.

Here are two recent examples from the current US election.

During the second Presidential Debate, Mitt Romney talked up some pro-women action he had taken as governor to get more women into the cabinet:

“We took a concerted effort to go out and find women who had backgrounds that could be qualified to become members of our cabinet. I went to a number of women’s groups and said: ‘Can you help us find folks?’ and they brought us whole binders full of women.”

That bizarre phrase “binders full of women” was picked up on Twitter where the hashtag #bindersfullofwomen became the top trending topic. Obama’s campaign jumped on the opportunity and bought a promoted tweet against the hashtag which linked to a donation page.

In fact the Obama campaign is clearly thinking in terms of hashtags. They are currently using the tag #romnesia to attack Romney’s tendency to switch positions on key issues. This tweet from Barack Obama’s Twitter account is typical:

“If you say you’ll protect a woman’s right to choose but you say you’d be ‘delighted’ to sign a law outlawing it, you’ve got #Romnesia.”

Social media will play its part, but the US election will also be fought out using the traditional heavy artillery of high frequency TV advertising.

Both campaigns have absolutely vast amounts to spend on TV slots. But there is a fascinating divide in the way that the media spend is actually being spent – which perhaps goes to the heart of the ideological differences between the two candidates.

According to an article in Politico, rather than using a specialist media buying agency “Romney does all his TV buying in-house through a lean operation headed by a single chief buyer.”

This is giving Romney a great deal of flexibility and granular control over buying – but he has missed out on a lot of efficiency gains by not pre-booking TV slots. The Politico article quotes the example of a single commercial break within Wheel of Fortune in which both Obama and Romney advertised. The Obama campaign paid $500 to reach Ohio voters: Romney paid very much more at $2,800 for a spot in the same break.

So this US election pits the skills of a dedicated in-house resource against the deep resources of a specialist agency.  For all those in the agency world bemoaning the tendency of clients to take work in house, this could turn out to be a massively thrilling or depressing election result.

While I can work on simple hypotheses like ‘if this is written by Melanie Phillips it is almost certainly ghastly rubbish’, a search engine like Google has to calculate authority through a dense mix of keyword density, social cues, link quality, time on page, speed of hitting the back button, etc, etc.

That’s hard work – which is why Google is now starting to account for authorship in its algorithm. It is identifying content by author, and is starting to calculate the authority of different authors. So all things being equal, if Delia Smith and Jamie Oliver are both writing about breakfast, the author with the highest authority will rank higher and (cue drumroll) bring home the bacon.

You have to go through a few hoops to get Google to recognise you as an author – and at this point I should pause to acknowledge the long-suffering natural search team at Harvest who have patiently answered my endless questions on this subject.

The path is hard, but the rewards are great. Here’s what you get:

1. A lovely email from Google welcoming you to ‘Google Authorship’.

2. Your very own photo now features in Google’s search results, terrifying small children and kittens who may happen across it.

3. Best of all, buried in the hyper-geeky labs section of Google Webmaster Tools, you now have access to your very own author statistics across all the web properties that take your work.

So how unbelievably cool is all that??

We’re entering a world where freelance journalists can argue up their fees based on their established authority – because higher authority means better search ranking, which ultimately means more page views and ad revenue for the publisher.

Of course it’s also a world where your employees can leave you and take their author authority with them.

So here’s my final point. The manager of Village People is meant to have had one of the easiest jobs in the music business. In any dispute with ’the talent’, he knew he could always replace the fireman or the red Indian with someone else in fancy dress and no one would be any the wiser.

Compare that with the miserable job of the manager of a ego-filled supergroup like the appalling Crosby, Stills, Nash and Young. Any one of the stars could flounce off leaving the brand irreparably damaged.

As I read it, the Village People model is now the one to avoid. Where authority is vested in authors, agencies and clients will need to recruit and retain their own experts – and perhaps work to promote their employees individual brands.

And ironically we have a model for exactly this kind of business in the ego-fuelled advertising agencies of yesteryear with their convoluted surname-rich names. Saatchi & Saatchi, Doyle Dane Bernbach, Ogilvy and Mather, Sterling Cooper Draper Pryce, Abbott Mead Vickers - you are not dinosaurs from another era, you are instead THE FUTURE!

Google Tag Manager plays nicely with Google and non-Google website tags. Key features include:

• Asynchronous tag loading — so your tags can fire faster without slowing down the visible part of your web page
• Easy-to-use tag templates, so marketers can quickly add tags with our web interface, as well as support for custom tags
• Error-prevention tools like Preview mode (so you can see proposed changes before implementing them), the Debug Console, and Version History to ensure that new tags won’t break your site
• User permissions and multi-account functionality to make it easy for large teams and agencies and clients to work together with appropriate levels of access

Harvest Digital is one of a small handful of UK agencies to be Google Certified Partners for Google Tag Manager – if you want to find out more, please drop us an email at info@harvestdigital.com.

A few weeks ago I blogged about Nate Elliott’s post on the Interactive Brand Ecosystem.

Essentially Nate is arguing that we need to move from a TV-centric way of thinking and adapt to a planning process that puts the website at the heart of digital marketing.

I agree with this, but I don’t think the model is completely right for direct response digital marketing activity. If you skip to the bottom of this post, I’ve had a shot at revising the model.  Having said that, the entire landscape right now is in a state of flux so it’s difficult to predict exactly where all the pieces will end up.

What does the Interactive Direct Marketing Ecosystem look like?

This is a fascinating time to be working in direct marketing. Over the past ten years, the rise of digital marketing has created an entirely new way to interact with customers.

In fact over that decade digital marketing has grown from virtually nothing to be the largest single marketing channel – bigger in marketing spend terms than press or TV. The change has been remarkable – and yet I believe the biggest changes within the channel are yet to come.

The vast majority of marketing spend on digital is in pursuit of direct marketing goals. In the UK, 80% of spend goes on paid search and online classified advertising which are essentially focused on the direct acquisition of customers.  Even within display media, a very small percentage of spend goes on the kind of interruptive, interactive formats that are used to support brand campaigns.

To show how things have changed and are about to change still more, I want to focus on three key areas for digital direct marketers – segmentation, permission and email marketing – and look at how they are changing.

1. Segmentation and targeting

The gold standard of offline targeting and customer segmentation is demographic targeting where target audiences are defined in terms of age, sex and socio-economic group.

Demographic targeting is also used in the digital world, but generally it plays a subsidiary role to behavioural targeting.  In fact much of what we do from a planning point of view can be summed up in a simple philosophy – don’t ask, observe.

At its purest form, choice of search terms is a brilliant indicator of purchase intention.  Search marketing then lets you overlay keyword selection with other factors like time of day, day of week, location and – belatedly – demographic information.

But demographic information can be a false friend. The office manager and the company director will have very different characteristics – yet the director’s fancy foreign trips may be being researched and booked by the office manager. Search marketing avoids these issues – if you are searching, you are potentially researching and purchasing.

Display media is where targeting gets really interesting. When you land on a web page, before a banner is served there is a ‘data exchange event’ – a real-time conversation between your web browser and multiple adservers to work out which banner should be displayed.

Whilst all this data is anonymous, it is still incredibly rich.  We can potentially target on some or all of these factors:

Demographic data – age, gender, income etc.
Environment – device, operating system, type of browser, bandwidth, ISP
Behaviour – recent visits to a website, recent search history
Context – content on the page where the banner is being displayed. Targeting could be by relevant section – say the finance section – or targeted to specific keywords appearing on a page.
Time and location – time of day, day of week, computer location.

To add a further level of complexity, some of these criteria could be used to target a specific audience and others could be used as triggers for dynamic creative.

For instance, Harvest Digital has created campaigns for Interflora using dynamic creative based on time of day.  If the banner is displayed before 3pm, the call to action is: “Order now for same day delivery”; after 3pm, the banner automatically switches to “Order now for next day delivery”.

To give another example, Orbitz in the United States realised that visitors to their website using Apple computers tended to have a higher disposable income. So when their banners are shown on Apple devices, they automatically show hotel rooms that are 30% more expensive than the norm.

The rise in web browsing from smart phones and other mobile devices is now giving us yet another segmentation opportunity – user location.  For instance, we can bid more to display ads on the mobile phones of consumers who are physically at certain locations – for instance, within walking distance of a particular retail outlet.

These are all exciting opportunities for direct marketers.  We have moved from an essentially static form of targeting based at its core on physical location – i.e. demographic data derived from postcodes – to fluid, real-time targeting based on the changing patterns of consumer behaviour.

2. Gaining permission

This is a really interesting area because change has for once been driven by legislation and consumer action.

Permission in the narrowest sense refers to explicit data capture – normally in digital marketing the acquisition of an email address. As I go on to argue, this kind of permission and the value exchange that supports it is more important than ever.

However, particularly in Europe, we are seeing the development of a secondary level of permission around the acceptance of cookies.

Marketers have generally seen this requirement to get some form of (in practise quite weak) permission to use cookies as a major inconvenience. However I’d suggest that it can be viewed in a positive light.

The ‘permission landscape’ now looks like this.  The inner circle is customers where you have transaction related emails alongside marketing email.  Beyond that is prospects who have given you permission to email them, perhaps in exchange for a software trial, a free offer or an informative newsletter.  And now there is an outer circle, of consumers who have visited our website and given us permission to deploy a range of tracking cookies.

This final category can be reached through retargeted display advertising. We can’t personalise communications since the cookie data is anonymous – but we can infer quite a lot from the other data associated with that cookie.

From a content point of view, retargeted advertising has got a little bogged down on simply replaying the last product you looked at on a website. That’s OK on an ecommerce site – although as the same fridge follows you round the Internet it does get tedious quite quickly.

However, there is much more that could potentially be done via retargeting. It is possible to drive retargeting strategy off rich web analytics data

3. Email marketing

Finally email marketing. Email is important to digital marketing as the primary channel we use to deliver CRM programmes – but the rise of mobile devices is turning email into a much more powerful channel.

According to MailChimp research, in the UK, some 30% of email is now delivered to mobile devices. In the US, that figure is over 40% and in Japan it is approaching 60%.

So email is no longer a way of reaching customers on a desktop computer during work hours – it is a way of reaching customers in real time to wherever they are.

This is already making profound changes to the way we create email marketing campaigns. Creative needs to be adapted to work on smaller screens operated by fatter fingers. Conventional thinking about email delivery times needs to be revised.

But beyond this, mobile email will need to adapt to the capabilities of its environment. We can see the possibilities by looking at mobile search marketing, which in addition to offering a conventional click-to-website functionality also offers click-to-call and click-to-view-map.

Thinking further into the future, the mobile phone will become the primary payment mechanism – a kind of mobile wallet. At that point, email becomes an extremely powerful channel to deliver vouchers and offers direct to the consumer, direct to their payment device.

The interactive direct response ecosystem

These add up to really important changes in areas that are the bedrock of the new direct marketing. So what does this do for our diagram?

I have left offline media off this diagram, because I just want to focus on the digital journey. The big changes here are identifying remarketing and email as separate channels.

Here’s another way of looking at these relationships, which shows how channel strategy lines up against a typical customer journey.

So where are we heading? Across targeting, gaining permission and email marketing, the potential is there for digital direct marketing to be even more effective than before.

However, it is vital that we bring consumers with us on this journey. We have seen from resistance to text marketing that consumers are very protective of what they see as personal spaces – and hence people may become even more resistant to email marketing as it moves from desktop to mobile.

The key to success will I think come down to the central concept of permission marketing – the value exchange between customer and company at the point at which permission is granted. For this to be a value exchange, we need to have something of value to offer – relevant content, great offers, exclusive information. If we don’t offer value, this will become a sterile conversation that consumers will turn their backs on.