Of Web Personalization Gnomes

A new entrant has this week joined the web personalization fray. They come in the form of UK based firm Magiq and I’ll warn you, they come armed with gnomes who will put your campaigns live and take care of the tech aspects!

If we put Magiq’s silly in-jokes to one site [just below the shrubbery? - ed] they appear to have quite a neat self-serve personalization platform that will allow marketers to include CRM data in their targeting of visitors on the web. In their own words:

“Magiq allows you to take immediate and real-time control of your website’s personalization, without needing to involve your IT team or hosting provider.

So whether you are the site’s in-house marketing team or an agency wanting to help your client’s on-line marketing, Magiq lets you set up Gnomes to power the campaigns immediately, and change them at will, responding to changing demands and opportunities without overloading IT or being stuck in endless change control processes.

If you’re IT then Magiq can be a great way to remove the workload of constant changes to the site generated by the marketing and CRM teams.”

Magiq web personalization

Interesting stuff. Case studies and client testimonials are at this time scarce on their site other than a Dutch insurer but there isn’t much detail on that one either. Anyone had experience with this solution and care to comment? From the postal address we suspect the same guys might be behind this web personalization solution as who drive the Speed-Trap analytics technology. Speed-trap are partnered with the likes of Teradata and SAS so presumably they can suck that data into their website personalisation solution. Tech details are scarce at the present time though.

Meanwhile, Maxymiser have again updated some of their documents and case studies on their web personalization solution this week.

MVT

Multivariate Testing Experiences at Glasses Direct

The guys at maxymiser have put out a testimonial from David Carruthers at Glasses Direct on his experience of using multivariate testing with them. This one is interesting as it’s a client who has had experience with a competitor. The posting doesn’t say who the other supplier was but it does say paid-for so we can deduce it wasn’t Google Website Optimizer.

Glasses Direct Multivariate testing image

Unlike many case studies, they’ve posted this on their blog so you don’t need to give your email address to get hold of it – take a look at the testimonial here.

In other news, SiteSpect continue their expansion in europe with their Netherlands office starting to make an impact. Omniture appear to have removed many of their multivariate testing case studies but there is still one from Readers Digest (a brand that went bust in the UK last month).

Multivariate testing is moving forward apace. some content will follow on segment targeting allowing markters to target visitors with different content depending upon where they have come from. If multivariate testing is new to you, hit that link and get the maxymiser explanation, one of the more scientific explanations in our experience, lacking the marketing fluff of the others.

Web Personalization – Good or Bad?

Since my last posting, a few people dropped me a note to suggest that I sould use US rather then European spelling, without wanting to give away my location, let’s settle on US spelling for now. From here on in, it’s web personalization. Other feedback from the article concerned the worries of users. Rest assured, all the vendors we’re covering consider web personalization to be something that makes the content of brand sites more relevant to the user.

We’re not talking Phorm style whole of web targeting based on ISP spying or even Criteo web personalization of general sites based on products browsed in an ecommerce website. We’re talking about a brand taking the behavioral data they hold about a user and using it to present more relevant and timely personalized offers on future visits.

Omniture does this a for a number of brands already, they acquired a product in about 2006 called Touch Clarity which is now branded Test & Target One to One. The idea is that it takes visitor level data and uses it to achieve a level of website personalization. Some find it’s quite a black box approach but black boxes have their advantages, they can be left to run and their algorithms will optimize automatically overnight or whilst you enjoy the weekend!

Other firms doing it include x+1 and also a european firm recently expanded into the USA called Maxymiser. It sounds a little like a tacky name (or snake oil sales person esque) but their web personalization is one of the few to extenstively document how it takes even offline data and uses that to make personal recommendations for products and content or offers.

We”ll take a look and review some of the offers shortly. At present all vendors are being quite tight lipped on actual demos, understandably so because they’re in the process of protecting their innovations with patents we assume.

M.V.T.

Web Personalisation comes of age

The natural progression of multivarite testing has often been said to be personalisation of web content based on on-site and off-site data. This includes CRM data. The potential is enormous, brands could unlock the vast amounts of data they hold in offline CRM from ‘real world’ transactions and behaviour and use that for web personalisation.

Your author has become a little excited because Unica just listed this very trend as number 4 in its 10 Key Online Marketing trends for 2010. Jay Henderson writes:

Marketers Expand Targeting and Personalization on Their Websites

Expect more personalization in websites in 2010. 55% of marketers are already using targeting and personalization on their site, and another 21% will roll it out this year. This personalization will grow more sophisticated as well, as marketers target “anonymous” visitors based on referring URL, search terms, geo-location and other insights.

Marketers will also extend email segments, offers and messages to websites, to better match customers with marketing messages. Landing pages and website optimization will complement site personalization to start delivering superior results.

Although I laud him for pushing the web personalisation cause, I’m not sure where the 55% of marketers personalising their websites comes from. Okay, that percentage probably greet their customers by name but web personalisation has to be something more than that doesn’t it? The author is an ex Unica VP by the look of things so should know his stuff.

Let’s hope brands take this to heart and get ready for real web personalization!

Landing Page Optimization – another name for Multivariate Testing?

We’re never quite sure whether landing page optimisation (or landing page optimization if you’re in the USA) should be considered a different discipline from multivariate testing. Certainly, the objective end of landing page optimization focusses pretty much totally on tools like a b and multivariate testing.

Landing page optimization can probably also be considered to include pre-testing type work such as web analytics analysis, heatmapping and other tools. User testing can also contribute to landing page improvements if in a more subjective way. Consultancy services can also be used but such opinions are probably open to more error and are highly subjective and based on gut-feeling more than scientific data.

Optimization is good however you tackle it. The key thing is to improve the conversion of your landing pages through some means. Even the google web optimizer tool is a good start point but most will become frustrated with its lack of support for multiple action tracking or more complex tests once the checkout and other pages become candidates for improvement.

To find out more, speak to an agency or one of the vendors in the space. Most vendors avoid the term because it’s pretty vague and covers a lot of things. Maxymiser recently put a page live in their new site on the subject, it doesn’t give too much by way of broad detail, calling landing page optimization a form of multivariate or a b testing really. Hit the link to take a look if you didn’t already spot how the web works!

Enjoy!

Multivariate Testing – The Changes Continue

There’s been a bit of a gap in postings, that is no sign that multivariate testing is any quieter, it’s just your writer has been busy. All players are still hard at work winning new clients and producing excellent restuls for some others.

Optimost have continued to win some interesting names in the UK and USA, most of their attention appears to be on upselling to existing Autonomy clients. Unfortunately many of the Optimost original team has parted from the company causing many to question the level of expertise left at this firm.

Maxymiser recently took part in the European Tech Tour of leading start-up hi-tech companies. In other updates, they’ve also signed Lastminute.com recently and ran a client seminar that had EasyRommate, bmibaby.com and Cancer Research UK speak in person. Slides haven’t been published but a gentle email to the guys there should get you a long way. Maxymiser is also pushing into the US market, there’s no US names on the site but we hear there are a couple of big names under NDA.

Omniture continues to hand the keys to multivariate testing (test & target) to all and sundry and doesn’t necessarily provide much support beyond that point. That said, it’s like Google Web Optimizer but better and better supported and for a good price so they’re building market share.

Meanwhile analytics vendor WebTrends continues to levearge its Optimize(tm) testing tool into existing clients.

The much promised update on the whole market and a vendor comparison will follow. Your author has now had hands on almost all major multivariate testing tools. There’s been no good indipendent reviews of all the tools yet, hopefully we’ll be able to offer the first for you. Comments welcome as usual!

ASDA in Multivariate Testing Success

The top UK supermarket ASDA (formerly part of the wal-mart family) has gone public with some great results of a multivariate test. For the full gen, see Internet Retailing magazine’s coverage here, there are a few excerpts below for your edificaiton.

A multivariate test to the ASDA homepage saw bounce rate reduce by 19% so that’s a bag-load more visitors going deep into the site and shopping in one of the ASDA brands (groceries, finance, whitegoods appear to be the major ones). There’s no comment on the impact on conversions at the end of the funnel but Chris Dalrymple of ASDA is very positive about the result in the quote:

“By making small and subtle changes to our website we have already seen that this can have a huge impact on our customers’ ability to find what they’re looking for. This is a level of detail that we’ve never previously had access to whereas now we can use Maxymiser’s analysis tool to look at changes that we know drive conversion uplift. ”

It’s great to see multivaraite testing getting some coverage in the vertical press over here in europe. So many of them have been blind to its impact to date even though vendors including Omniture and Optimost have been plugging away and publishing some great results. Perhaps the press are so used to amazing stories of uplift that they are already jaded when they hear of this new multivariate testing thing?

There is also apparently a case study pdf of the results on the Maxymiser website but I have to rush so will leave you to locate that!

Maxymiser Increase Conversion Rate Challenge

Maxymiser today blogged about this challenge they have laid down. They’re intersted to know whether you can guess the winning content from an a/b/c… test with four variants. They have disguised the client’s variants for annonymity so it looks fairly generic.

I had a go and was quite surprised by the result, 2 guesses was what it took me, what will it take you? According to Maxymiser’s data, the result of this test led to an 11% increase conversion rate for their client. Looking at other material on their site, the client might have been UK DIY store chain Wickes (Hardware store if you’re in the states).

Looks like the general maket is picking up on multivariate testing as a reiliable way to increase conversion rate, it’s not yet fully mainstream but certainly, all the early adopters are having a tweak and are having some great results. Large multinationals should be next on board, Google’s results with YouTube were great last week, will post the link when I find it!

Anyway, you were probably waiting for the link to the Maxymiser challenge – here’s the increase conversion rate challenge, let us know how you do.

Update on all things multivaraite testing related

It’s been a while since I posted so here’s a rundown of what I missed in the multivaraite testing space.

Omniture have poached a leading corp dev person from Interwoven/Autonomy/Optimost. No doubt he will help that company proceed with their highly agressive acquisitions plan.

Autonomy Interwoven Optimost has launched what it terms “Meaning Based Marketing“ but I and many I talk to are struggling to fathom what it all means. Is there anything new other than the ability to target content to visitors dependant on some on and off site factors? Is it just a multivariate testing platform with some marketing varnish?

Maxymiser has appointed Tim Brown, ex-CEO of Touch Clarity as CEO. The multivariate testing, targeting and behavioural targeting firm hopes to expand geographically with Tim at the helm.

That’s all for now, expect more as it happens. Updates will be more frequent in future, back on stream now!

Another Multivariate Testing Case Study

Maxymiser appear to be leading in terms of case studies on their results. In the early days, other multivariate testing vendors such as Omniture and Optimost came out with the best case studies but these guys have just released some good results with UK dating site LoopyLove who have been using multivariate testing to optimise their landing pages and sign up pages.

LoopyLove winning and default variants from their recent multivariate test

To borrow a quote from the case study:

“Loopylove engages Maxymiser to improve site performance, continually testing and improving web content. This test delivered them a 31% increase in membership signups.”

We’ve pinched the pictures of the winning and default variants from this multivariate test for your viewing pleasure. Case studies with all the variants trialled in between are rare, I guess clients and vendors alike consider those to be commercially sensitive.

By rearranging the form options on this first page of the member sign up process and redesigning the progress indicator, LoopyLove have increased conversion to membership by more than 30% according to Maxymiser’s multivariate testing case study on the subject.

It goes to show the dramatic effect that subtle changes to content can have on visitor conversion rate. Multivariate testing is a great means by which to measure that effect and to ensure that your changes are producing uplift rather than down-lift. LoopyLove have learnt valuable lessons here that will allow them to better convert traffic for many years in to the future.

Or should we call them ‘client success stories’ as Maxymiser now does. I guess a case study sounds a little dull so perhaps that new american-style moniker is an improvement, what do you think?