Social media measurement is something that I think should be undertaken with a sense of perspective, by standing back and looking at the big picture.
A widescreen approach to social media measurement ultimately looks at the things that really matter: sales, profits, customer satisfaction and loyalty. Besides, honing in on the detail might not be the best use of your time, given the obvious difficulties that arise, particularly with attribution.
But standing back and looking at the bigger picture is not going to be enough for your data-mad boss, is it? It’s a bit too soft focus, right? He or she is going to want to see proof that all this social optimisation is actually working.
If that’s the case, then don’t worry: there are lots of things you can measure…
It’s all about engagement
When we talk about social optimisation (a term I prefer to ‘social media’) we’re really talking about driving engagement and interaction. The goal of any social optimisation strategy is to provide the right tools so that people can engage with your brand / people / products / services onsite and offsite.
Here’s what you want to happen:
- You want people to make a noise.
- You want people to store and share things.
- You want people to love your website.
- You want people to visit more frequently
- You want people to refer your company to their friends.
- You want people to buy into your brand.
- You want people to buy your products.
Engaged customers and prospects are far more likely to do some or all of the above. So how can you boost customer engagement?
Give people the right tools
The tools and onsite functionality you need is going to depend on your business, your strategy and your goals. What you’re ultimately looking for is a wide range of tools to help people interact. It doesn’t matter whether this interaction happens onsite or offsite, but only that it happens. You can measure it either way.
This list of KPIs / metrics should help you figure out what can be measured (at a nano level) and also what kind of tools / functionality you may want to introduce. I still think it’s best to measure from a distance but if your boss wants the detail then this list should help you work out what to look at. In doing so you’ll able to determine the relative success and adoption of new features. You may also unearth trends and spot opportunities or issues.
In any event, taking a top-down look at interaction - and monitoring how customer engagement changes over time - can really help you position your company as a community-centric organisation, by proving that an investment into customer engagement is a wise one. Your boss should be happy if all goes to plan.
Making interaction a game
This list has been largely informed by a new social commerce startup that I’m working on. It’s essentially a marketplace that connects buyers with sellers. I created a ‘kudos’ algorithm that helps us curate the website. Items that are highly rated and that attract lots of comments / bookmarks / followers will gain kudos points. We apply different weightings to different interactions (for example, a ‘love this’ rating is worth less than a ‘follow item’). Editors / curators can then spot the buzz and act accordingly (better promotion, interviews, videos, etc).
We created ‘kudos’ for a few reasons. Firstly, we want to learn from the crowd. Secondly, we want the website to be highly interactive. Thirdly, we want it to feel like a game for the sellers, just like Digg is for the article submitters.
So tracking and making sense of interaction is a fundamental part of our web venture. Many of these metrics are factored into our algorithm, and in the same way you can score different interactions to create some kind of interaction index. It might help you condense all of this data noise into a more digestible format.
Before we jump into the list there are a few caveats…
- Not all websites are equal. Not all of these will be relevant to all sites (e.g. ‘Posts’ won’t be any good for sites without blogs and contributors)
- Not all interactions are equal. ‘Print page’ as an engagement measure is barely worth looking at… or is it? In any case, some of these things are more important than others (hence my scoring / ‘kudos’ algorithm).
- There is some crossover. For example ‘bookmarks’ and ‘wishlists’ may be the same thing on your site (although they’re not on mine).
- Some metrics will have sub-metrics.
- Avoid curve balls. If the widget sucks then it doesn’t matter that 10,000 people installed it last week. It will still suck and they’ll hate it.
- Human power is needed to really understand the detail behind the numbers, and to act on that knowledge. Interpretation is key.
- It’s about quality not quantity. Don’t go counting those spam comments!
- This is a bit of a braindump and I’ll certainly have missed out various things, so please leave your pointers and suggestions in the comments section below. What are you measuring?
A list of social interaction metrics / KPIs
- Alerts (register and response rates / by channel / CTR / post click activity)
- Bookmarks (onsite, offsite)
- Email subscriptions
- Fans (become a fan of something / someone)
- Favourites (add an item to favourites)
- Feedback (via the site)
- Followers (follow something / someone)
- Forward to a friend
- Groups (create / join / total number of groups / group activity)
- Install widget (on a blog page, Facebook, etc)
- Invite / Refer (a friend)
- Key page activity (post-activity)
- Love / Like this (a simpler form of rating something)
- Messaging (onsite)
- Personalisation (pages, display, theme)
- Profile (e.g. update avatar, bio, links, email, customisation, etc)
- Print page
- Registered users (new / total / active / dormant / churn)
- Report spam / abuse
- Social media sharing / participation (activity on key social media sites, e.g. Facebook, Twitter, Digg, etc)
- Tagging (user-generated metadata)
- Time spent on key pages
- Time spent on site (by source / by entry page)
- Total contributors (and % active contributors)
- Uploads (add an item, e.g. articles, links, images, videos)
- Views (videos, ads, rich images)
- Widgets (number of new widgets users / embedded widgets)
- Wishlists (save an item to wishlist)
Any good? Rubbish? Let me know what you think, and what I missed…