Are you an AMA member? please login
19th April 2019 Jacqueline Haxton

To takeover or not to takeover? #DigiLab

That is the question that Digital Lab Fellow Rachael Williams explores in her second #DigiLab blog post as she considers the pros and cons of website homepage takeovers as a new website functionality allows the British Library to further test how its users like to consume key brand messages and launches.

Homepage takeovers are a big statement for a brand or company. What do they want to shout about so much that they’ll give it pride of place on their homepage — often overruling all other messaging available to users landing on the page? A big decision when your organisation serves a multitude of different users trying to reach very different outcomes. In our case, over one million people use our website every year to do things like research their family history, find primary resource material on typographical techniques in Germany, seek out tips for starting their own hair care business, peruse our shop’s stationery offerings, or buy tickets for a blockbuster exhibition.

In short: we have a lot of users to guide and content to provide.

At the request of our Head of Digital Content Strategy, our Technology team developed the functionality for our brand to showcase homepage takeovers. Once ready, it was down to us to test what type of content was to be served and find out more about what our web visitors reacted well (and not so well) to.

A quick look at our everyday www.bl.uk homepage — our events and exhibitions are featured prominently, as is the ability to search the Library’s catalogue. Scroll past the fold and you’ll see news items, blog, ways to join us, featured collection items and a shop carousel.

The takeovers give prominence to one key cultural/marketing message but retains the top drop-down menus and the search the ‘Main Catalogue’ option (as many users will be coming to our site solely for research purposes).

Over a five-week period across October and November 2018 we tested a number of content types in the homepage takeover platform. We chose content that served Library values and purposes and that would help to deliver KPIs (key performance indicators) such as driving traffic to web spaces (curated micro-websites on specific subject areas), increasing the number of digitally consulted collection items, and selling tickets to exhibitions.

Varying content type, imagery and calls to action, we served ‘takeovers’ for a maximum of seven days — for our major exhibition Anglo-Saxon Kingdoms — and three days for all other tested content, to ensure we could measure results and behaviour for a substantial period of time.

Some other examples
Following the success of our Windrush exhibition we wanted to promote the launch of its accompanying website — Windrush Stories — using the same striking lead artwork:


 Do big names sell? Our launch of the Winter Events featured some of the key speakers taking part in the season ahead:

Cats and the internet. The purrfect combination. Paw prints prowled onto the homepage for the launch of our free Cats on the Page exhibition:

We also ran tests linking to our World War One webspace for the centenary weekend of the Armistice.

My hypothesis, based on user behaviour email, was that a playful or unexpected call to action would get more clicks than a ‘Book now’ — as (it’s thought) that people are put off by the idea of parting with their money so quickly. Using my awareness of how posts have gone down with our social media users, I also thought a purely events-focused message wouldn’t be as popular as other content, such as free web spaces that offer users a wealth of amazing collection items and stories to explore.

The results: a snapshot

  • Anglo-Saxon Kingdoms was the top-performing takeover, receiving 3,487 total number of clicks on the call to action (CTA) button over seven days (avg. 489 per day), followed by Winter Events, which received an average of 291 clicks on the CTA per day. Windrush Stories was the least successful, receiving 325 CTA clicks over three days.
  • All of the takeovers performed better in terms of driving revenue from the homepage than the same number of days in the previous week.
  • The best takeover topic in terms of clicks, click-through rate and revenue was Anglo-Saxons. Even though it ran for seven days rather than three days for the other takeovers the daily metrics are still greater.
  • ‘Book now’ is the strongest call to action used out of the five — interesting as we tend to steer away from such a definite CTA in email so as not to ‘scare’ people off.

So without making things sound too essay-like
Here’s my analysis, or at least food for thought, from this experiment so far:

  • Our World War One web space is one of our top-performing microsites so I would have expected more people to click-through. Were some users confused seeing this on the homepage? Unsure of what they were being delivered so abandoned rather than pursued?
  • Expect the unexpected: Windrush was one of our most-visited free exhibitions ever, we should try to play with text and calls to action if we feature this on the homepage again. We know there’s an appetite for this content (as the exhibition visitor figures demonstrate) and just need to think of better ways to serve to get those clicks up.
  • The diversity of our audience should never be underestimated. Our Anglo-Saxons exhibition has attracted over 90,000 visitors and the reaction to seeing this on the homepage was astonishing; this exhibition appeals to a very Library-loyal audience. But we should be experimental too — Cats on the Page is playful, bold and different to our ‘expected’ themes and material, and if our social media followers are anything to go by, they love moggies on- and offline
  • Our Anglo-Saxons takeover was the first one we tested — were people especially curious of this new landing page experience which helped to generate the very high click rate?
  • While successes like a spike in ticket sales may be the result of a takeover, we should continue to monitor the impact on other areas of the website that are reliant on homepage traffic, i.e. blogs and vacancies. Customer Services can also provide us with qualitative feedback from users who have landed on a homepage they were not expecting (and perhaps didn’t like).
  • Are our users bolder and braver than we think? They don’t seem to have been put off by clicking ‘Book now’.
  • Higher click rates on the ‘takeovers’ may be due to the design itself rather than content alone. As mentioned in the beginning of this blog, we have a lot of users to serve and our homepage tries to tick a lot of boxes when it comes to what people are looking for. This bolder, more prominent ‘singular choice’ approach seems to have a positive effect. Are people deterred by too many options on our standard homepage?
  • Short bursts of carefully considered takeover content is working. Our team have drawn up criteria for proposals for homepage takeovers, this will evolve as we find out more about our users and their reaction to different types of content.
  • How do the takeovers effect user behaviour otherwise I hear you cry (or should that be meow?)? As expected, unique page views on the content being directed to is significantly increased (around 80% more) the days of a takeover, compared to the same days the following week. Interestingly however, the dwell time and bounce rate on those landing pages are not as dissimilar when comparing corresponding days the week after a takeover. Obviously with something like an exhibition, a spike in ticket sales is no doubt down to more people going to that web page — leading them on a journey to purchase – all stemming from the start point of a homepage takeover. But with takeovers where we’re not ‘selling’ it’s hard to fully interpret the results at this early stage. I guess just because we’re putting the content more predominantly in front of someone, they still may not be wholly interested in it. Maybe if we time certain content types at more opportune moments, we’ll capture people’s attention — and hold on to it — a little better. Which leads me on to …

Next steps?
We have some exciting exhibitions and webspaces launching over the coming months which will be given homepage takeover slots. It’s likely that we’ll also re-try events season launches and exhibitions with longer runs, so we can compare how Cats performs at ‘Final weeks’ push with its opening message. With LGBT History Month and International Women’s Day in February and March, it would also be topical for us to try more ‘timely’ takeovers by giving our LGBTQ histories and Sisterhood and After webspaces prominence on the homepage.

We shall continue to test these mini experiments and learn from the results and I shall purr-sistently try to weave cat puns in to my work at every opportunity for the duration of our Cats on the Page exhibition.

Pawsome.

Read Rachael’s first #DigiLab blog Finding that golden nugget among a hoard of treasure #DigiLab

Images courtesy of the British Library ©.