I was at the Connected Local Government Live Conference last week and at one point the issue of organisational need versus user need came up.
Someone said that their users clearly needed to know about the events the council ran, because when those events where put on the homepage of the site, the engagement in the events went up tenfold (I think numbers like 30,000 were mentioned).
While this does seem to show that people want to engage in events, and that putting them on the homepage makes people aware of them, I think it misses some fundamental details of how council websites work for the people using them, and it shows how only measuring half a transaction can give you false results.
If you put something on the homepage, people will view it; this is undoubtedly true. Even when people search Google for what they want and often go straight to the page in your site where that is, your homepage is likely to still be the most visited page on the site. Putting something on the homepage will increase the number of people who see it and click through, even if that thing isn’t what they came to do. It’s just more likely to happen.
Lots of councils used to - and still do - put news front and centre on their homepage, and quite a few people will click through to the news from the homepage. Typically, you see a 3%-5% click-through to news from the homepage of a council site (interestingly - that doesn’t change much if the news takes up the whole center of the page or a bit near the bottom).
Just looking at the raw numbers of visits to things on the homepage versus those that are not will make it look like those things matter to your users; but if you do that you are only measuring a small bit of what’s going on.
Most1 people come to a council site to do a thing; be it pay a fine, work out when to put their bins out, or find the local library. Very few come to see what it going on.
This means that the majority of people visiting your homepage haven’t come for the news or event info. Yes, you can distract them from their task by flashing things up when they visit, but - and this is the important thing - how much does that distract and inhibit people from completing the task they came to do?
Every bit of homepage real estate that you move away from helping people achieve the thing they came to do, will increase the failure rate of people in doing those tasks. Failures on your website will not only cost the person visiting the site, they will also cost you. Survey data shows that around 45% of people who fail at finding the information they want online will then phone the council.
If, for example, your homepage gets 10,000 visits a day, then if you make it 1%2 harder for people to find the thing they want, you will get 100 failures and 45 extra phone calls. Every. Single. Day.
If phone calls cost the council £2.70, and the transaction on the web costs around 20p3, that’s £112 per day for every percentage point harder it now is to find things on the homepage. Over a year that’s a cost of more than £40,000 and this doesn’t take into account any knock on costs, because when people don’t find what they want once they are less likely to try again next time. £40,000 is someone’s job. Maybe yours. Or a librarian. Or a social care worker.
Now, while these numbers are not exact, and there can be many factors to make the calls cheaper or more expensive, it shows you that there is a cost to changing pages (especially the homepage) of the site even a small amount.
It might be true that as an organisation the council is willing to make these choices, where they are willing to sacrifice the cost savings for the promotion, but the impact of these decisions across the whole council needs to be understood before they are taken.
1 most is a scientific number, but i have seen may a survey where this is quantified, news, events and consultations, always score low when you ask people why they visit the council site.
2 You are also fooling yourself if you think a full on advert will only have a 1% impact on the pages usability.
3 these numbers are indicative, and based in part on SOCITM data, costs are a lot less generic and very based on services - but they give an indication of scale