Category Archives: Uncategorized

Free News and Faster Horses

We are going through a once in a lifetime shift in the media business, an inexorable move away from the economics that sustained ‘Big Media’ for the majority of the 20th century. From Chris Anderson’s Freemium to The New York Times erecting pay walls, the issue of paid content is about as hot as it gets right now. The underlying notion that you give the content to readers and pay for it through advertisers is going away, to be replaced with a plethora of different ideas about how it may (or may not) be possible to monetise content online.

All this in the face of multiple studies that consistently show only a minority of customers say they want to pay for content online:

  • PCUK/Harris Poll (5% of 1,888 UK adults said they would pay if their favourite online newspaper began charging).
  • Gfk (total 18% of UK adults in international survey of 16,800 said they didn’t want to pay for “content”, ie. “news,
  • entertainment and information sites such as Wikipedia”).
  • Continental (total 37% of 500 UK adults said they would pay micropayment, larger fee or monthly/annual sub for online newspaper/mag).
  • Olswang/YouGov (total 19% of 1,013 UK adults and 536 teens said they would make micropayments frequently, a subscription or otherwise pay for news articles online, on mobile or ereaders if there was no free alternative).
  • Oliver and Ohlbaum (“15 to 20% of respondents [survey of 2,600 UK consumers] said they would pay £2 a month for their favourite news website if it was the only one that charged”).
  • Forrester (total 19% of 4,711 US consumers said they would make micropayment, pay a sub or buy a bundled print/web/mobile package for online newspaper).
  • Boston Consulting Group (48% of 5,083 regular internet users in nine countries, including 506 in UK, said they would pay for online news).
  • KPMG (11% of 1,037 people aged 16 and over “currently spend anything on online media” – findings vary for different media types).

Taking all eight studies in to account, the average proportion of consumers who would pay for online content is 21.8%.

While this is interesting, it is pretty well documented that people are poor at understanding what they want. It surprises me therefore that we continue to treat these numbers as if they actually mean something about whether people will pay for content online, or whether there are workable models for paid content.

We need to start looking at these surveys with a healthy dose of scepticism.

While Rupert Murdoch seems to be treated as a sort of mad uncle of the internet currently with his “crazy ideas” to make people pay for content, in reality what he understands is that it is not a case of customers never being prepared to pay, just that they’d rather not. Sky cracked the subscription TV market in the UK with the introduction of Premiership football.

We seem to be treating a mass media meltdown as a certainty, when in fact there are still opportunities for the big players to make their existing models work – with a few tweaks. What if Murdoch included a subscription to The Times with Sky TV packages for an extra £1 per week? Give customers a workable concept and they may just sign up. Ask them to come up with the concept themselves, and they will be lost.

I would suggest checking out Bad Science by Ben Goldacre for further demonstration of how what people say is often miles away from what they actually think, or may think given different options.

A final example, and this is my favourite, that perfectly encapsulates how consumers have difficulty in really getting to grips with what they want, comes from Henry Ford, who created the mass market for automobiles. He was once asked about the importance of the ‘voice of the customer’ in his business. His reply was surprising: “If I’d asked people what they wanted, they’d have said ‘faster horses’.”

There is always opportunity for a person or organisation with strategic vision to come up with an idea that changes the game. Given a different context, people will have hugely different ideas about what they want or need.

In a world of horses, people don’t want cars. They want faster horses.

In a world of free news, people don’t want to pay.

The customer is not always right. Or better put, the customer does not always have the imagination to be right.


Volume Forecasts on Google Insights for Search

Google appears to have added a forecast function to Insights for Search, allowing predictions of search volumes up to 12 months ahead. See the query here for “bonds”.

There is already some guidance in the Google help pages as to the background calculations involved:

Insights for Search examines the past values for the terms you’ve entered. Based on those values, it extrapolates the future values, creating a forecast of search trends for those terms.

Obviously this is a useful additional function, however as the prediction model doesn’t take into account the context of the search term, or account for any business cycles that may be driving a specific market, these numbers may prove to be misleading at best.

There could be some interesting applications of this, certainly in terms of understanding Google’s view on the direction of either particular brands or keyword markets. Beyond that it will be interesting to track over the next few months the accuracy of any predictions made.

Google and Trademarks

Having run into some legal problems recently with clients regarding advertising against competitor trademarks on Google, I wanted to put together a post on this issue, and the current state of play. It seems pretty clear to me that globally speaking, no case has really got to the bottom of the particular problems associated with advertising against a trademarked keyword, or using a trademark in ad text on a search engine.

In terms of context, AdWords is Google’s flagship advertising product and main source of revenue, offering pay per click advertising, and site targeted advertising for both text and banner ads. Advertisers create ads and choose keywords, which are words or phrases related to their business so that when people search on Google using one of those keywords, the advertiser’s ad may appear next to the search results.

A company can choose a trademark protected term as keyword. For example, if you enter “nike” in the Google search box, you will get listings from advertisers who paid for placement with this keyword. Google puts those listings off to the right side of the screen, clearly marked as ads.

Google started allowing advertisers to bid on a wide variety of search terms in 2004 in the US and Canada, including the trademarks of their competitors and in May 2008 expanded this policy to the UK and Ireland.

Certain trademark owners have become concerned about their brand awareness and claimed that this behaviour violates their trademarks. Google came consequently under fire both in the media and from a legal perspective for allowing advertisers to bid on trademarked keywords.

In its 2008 Annual Report, Google commented on the lawsuits as follows:

Companies have filed trademark infringement and related claims against us over the display of ads in response to user queries that include trademark terms. The outcomes of these lawsuits have differed from jurisdiction to jurisdiction. We currently have three cases pending at the European Court of Justice, which will address questions regarding whether advertisers and search engines can be held liable for use of trademarked terms in keyword advertising. We are litigating, or have recently litigated similar issues in other cases, in the U.S., Australia, Austria, Brazil, China, France, Germany, Israel, and Italy.

Courts in all these countries ahve differed in their verdicts and reasoning oncerning the potential liability of search engines, and there is still no certainty as to whether Google will be viewed as an active trademark infringer when third parties make an improper use of trademarked terms on AdWords or whether the liability passes to the third party advertiser.

Certain recent decisions by the Court of Appeal in Paris were favourable to trademark owners such as Louis Vuitton, although this is based on a history of interpretation of trademark laws in France.  According to the Court of Nanterre Google is an active trademark infringer. The mere fact of suggesting the infringement by using the mark as proposed keyword is enough. On the contrary, the court of Paris took the view that Google does not use the trademark for identical or similar products or services in a commercial manner. So there is no trademark infringement. But nevertheless Google’s conduct was not deemed legal. Google’s liability was based on the common civil principle of fault due to the lack of preliminary control to check whether chosen keywords do infringe third party rights. The Court of Strasbourg, took into account technical measures implemented by Google (a filter and links to check third parties’ rights) and excluded Google’s liability on all grounds. This decision appears however to be isolated.

More recently, Google was held liable by several courts like the Cour d’Appel of Paris in the Louis Vuitton case but Google decided to appeal the case in front of the Cour de Cassation and the proceeding is pending. Google has now lodged an appeal before the French Supreme Court and the court has in turn referred the case to the European level by sending a reference for preliminary ruling to the European Court of Justice. Courts in other jurisdictions have also asked substantially similar question to the European Court of Justice, including the Dutch courts in December 2008 and the German courts in January 2009.

Such reference for preliminary ruling, as referred to in Googles statement, has not yet been answered by the European court and it is difficult to know when the judgment will take place and whether it will help to clarify the debate. Based on several industry sources, it seems there is some heavy lobbying on the Court of Justice including from the European Commission to confirm that there is no trademark infringement from Google and that it should be free to continue its paid search activity going forward as conducted so far.

In addition to the ECJ response, it is worth looking at other decisions pending in certain jurisdictions. By way of example, a US appeals court ruled in April 2009 that a trademark infringement lawsuit against Google by Rescuecom be sent back to a lower court for trial. The Court of Appeals for the Second Circuit in New York ruled that Google should answer to the claim that it infringed upon Rescuecom’s trademarks by selling AdWords ads to Rescuecom’s competitors which appear when Google users search for phrases containing “rescuecom”. Prior to this ruling, it was believed that this issue had basically been resolved by the courts in the US in Google’s favour; there are other cases which provided some precedent for the notion that third parties like Google weren’t liable in cases like this. But that is no longer a given, and Google will have to defend itself against this trademark owner’s claim.

Another interesting case is currently pending in the UK involving Interflora vs. Marks & Spencer. The world’s largest flower delivery firm has sued Marks and Spencer at the High Court in London for sponsoring the word ‘Interflora’ as a search engine keyword. The case could be an important test of how UK trade mark laws apply to keyword advertising. Interflora’s contention is that when a user searches for the word “interfl0ra” a specific request is being made for that trademark only.

There is some UK precedent for this position. The main UK authority on keyword advertising is the Mr Spicy case in which the trade marks owner sued Yahoo! for allowing Sainsbury’s to bid on the keyword “spicy”. The court summarily dismissed the claim and found that the only use of the trade mark itself was by the search engine user. The ruling went further and opined that even if there had been use by the search engine, the use would not affect the essential function of the trade mark as an indication of origin. However the significance of this ruling should not be overstated, because “spicy”is a purely descriptive term and Sainsbury’s therefore had a legitimate right to bid on that word.

The case of Reed v Reed concerned the comparable issue of the use of a trade mark as a metatag within a website (metatags are embedded keywords within a website that are picked up by search engines during the searching process). It was found that such use was not use for the purposes of trade mark infringement, and further was commented that the web-using public are well aware that search results contain irrelevant hits and sponsored links. He thought they would not be confused and infer a trading connection simply because their search returned such a sponsored link.

Although Interflora does not need to establish that the public is likely to be confused by Marks & Spencer’s actions in order to succeed, it has brought the issue of confusion into play by arguing that, although the Marks & Spencer sponsored link itself does not visibly display the Interflora mark, Google users will infer a trading connection that is detrimental to the distinctiveness of its trade marks.

Below are listed some links to relevant articles concerning use of trademarks in search advertising online, and legal cases pertaining to this issue.

5 Scientifically Proven Ways to be Persuasive in PPC

I have recently been reading Yes! 50 Scientifically Proven Ways to Be Persuasive by Noah Goldstein, Steve Martin and Robert Cialdini, a pop psychology book in which a huge amount of research on psychology and human behaviour is distilled into one readable volume.

The essential premise of the book is that persuasion is a science, not an art, and hence there are systematic ways of bringing people around to your point of view.

This really made me think about Google and PPC advertising. Is there another advertising arena in which the advertiser (or copywriter) has less space in which to be persuasive? To communicate the core message, or brand promise?

In a PPC ad there are just 95 characters in which to sell a brand, product or service, to persuade people that your ad amongst all the others on the page deserves further investigation. Yet despite the brevity of the message, this channel currently accounts for a huge proportion, maybe 50%, of all advertising spend. To badly paraphrase Winston Churchill: “Never was so much owed by so many to so few (characters)”.

So, while a PPC ad is at heart initially a creative endeavour, I wondered if there were ways in which the principles laid out in the book could be applied up front.

Although any learnings would be fed back and cross referenced with wider creative ideas surrounding the brand, product or service, the ultimate success of any individual PPC ad is judged on a purely scientific basis. Which ad has the best CTR or highest conversion rate?

So, here are my thoughts on some basic principles from the book applied to PPC:

1. Inconvenience the audience by creating an impression of product scarcity. In studies a simple change from “Call now, we are standing by” to “If the line is busy, call again” will greatly improve call volume by creating the impression that everybody else is trying to buy the same product.

  • Learning: Users on search engine results pages respond to an impression of scarcity.
  • Example: “Limited webinar places available. Sign up today”

2. Introduce herd effect in personalised form. A hotel bathroom sign informed guests that many prior guests chose to be environmentally friendly by recycling their towels. However, when the message mentioned that majority of the guests who stayed in the specific room chose to reuse their towels, towel recycling jumped 33%, even though the message was largely the same.

  • Learning: Use local messaging where possible by geo-targeting campaigns.
  • Example: “Join London’s biggest community”

3. “Because” makes any explanation rational. In a queue for a copy machine, a researcher asked to jump the line by presenting a reason “Can I jump the line, because I am in a rush?” 94% of people complied.  A reasonable excuse? However, even changing the reason to “Can I jump the line because I need to make copies?” (the same reason everyone is in the line) led to a 93% success rate. A request without “because” in it generated only 24% compliance.

  • Learning: Include “because” in text ads.
  • Example: “Because we offer a flexible service to fit your finances”

4. Ad campaigns that do not incorporate brands tend to not be remembered. There are numerous examples of ad campaigns that achieve increased sales for competitors as a result of unclear brand messaging. Split testing research on text ads hasn recently shown increased conversion rates and better CPA on ads that include multiple brand mentions, alongside core brand messaging.

  • Learning: Use the brand name, and not just in the URL.

5. Tired people tend to be more receptive to arguments. Two groups were presented a product demo, and then asked to evaluate the possibility of buying it. Group A was tired and a bit sleep deprived, group B was in good physical condition. Group A was much more prone to buy. Website conversion stats for many direct response customers reflect this trend, and it is common to see better conversion rates in the early hours.

  • Learning: Check website conversion stats and consider up weighting spend in the evenings.

Beyond these few examples, I would urge any marketer to read the remainder of the book. It is extremely well written, and the authors present in detail the general principles of persuasion and discuss an abundance of specific uses in engaing fashion. More than anything, they advocate the idea that you can and should test. Something close to the heart of any PPC marketer.

Yes! 50 Scientifically Proven Ways to Be Persuasive

Wolfram Alpha on Channel 4 news

I was lucky enough to appear in a Channel 4 News package on Wolfram Alpha last Friday 15 May.

It was a great experience, and a fascinating insight into how TV news works in 2009. My 15 seconds of comment was cut down from a 10 minute interview with Ben Cohen, the Channel 4 Technology presenter.

The piece as a whole is very interesting, and it is great that Wolfram Alpha is getting such a huge amount of mainstream attention. For what it’s worth, I think Wolfram Alpha has been rather overhyped, although it is clearly a fascinating project with some big implications for all imformation providers – Google, Wikipedia and anyone else in the content curation game.

Full Wolfram Alpha TV slot and extended interview with Conrad Wolfram

Google Adwords Bid Simulator beta release

Google has just introduced a new beta Bid Sinulator feature on selected Adwords accounts.

Using data from the past seven days, the bid simulator calculates where AdWords would have placed an ad at a different maximum CPC, how many clicks an ad would have received at those simulated positions and how much those clicks would have cost.

Bid Simulator can help answer questions such as “How many clicks could I get if I used a different maximum CPC bid?” or “What would my ad position be with a different bid?”

While it can’t predict the future, the bid simulator allows you to explore what could have happened if you had set different keyword-level bids.

The feature provides increased transparency into the AdWords auction and allows some more informed bidding decisions, all positive thing sin terms of running a better optimised paid search campaign.

Will be interesting to see if this feature is rolled out on all accounts, and beyond this, how robust the data proves to be.

Google Adwords Bid Simulator Reference Guide [PDF]

Leaked FT strategy document points to the future of online news

Cutting a print edition, appointing dedicated “news integrators” and a properly articulated online news presence are all part of the Financial Times’s “Newsroom 2009” plan, detailed in a leaked strategy document which has found its way online.

The document sets out a new publishing model designed to unite its cross-media processes, and ensure journalists put online at the front end of the editorial chain in every case:

  • Reporters are expected to add links, company data, write headlines and check for length on their stories, plus build on them once online.
  • Editing desks then check everything, link the content to ready-made newspaper subbing templates, before the subs desk finish the cycle and press the publish button.

Further emphasizing online as the place to break news is the decision to replace the third printed edition in London with a later second edition.

This does appear to be a genuine web-first publishing strategy, and a step forward from the FT’s current model which holds back some stories for its print editions.

It’s worth noting that the FT already does the two things digital refuseniks are always insisting will save newspapers: they charge a subscription and they focus on quality content. But the point is that the future of newspapers requires greater change than that.

It’s pleasing to note that the FT is really getting involved in Newsroom 2009 and beyond.