Hootsuite have just published a very interesting report on social media in 2021, based on a survey of nearly 12k social media marketers. Our two favourite bits in it are the advice about using social to replace in-person customer experience during covid and about talking to boomers on social.

What follows is a Q and A between Hootsuite’s Sarah Dawley and our Grace Kite about the findings on social media evaluation.

HS: Over 55% of respondents in our 2021 Social Trends survey don’t use an attribution model to prove the ROI of social media. (Last year this number was even higher, at 73%.) Does this surprise you? Why or why not?

GK: Small businesses make up the vast majority of advertisers on social media channels, and these can often see the impact of social even without tracking or modelling. So, no, its not hugely surprising to me. I’d be surprised if the advertisers that spend a lot weren’t using some sort of evaluation, though.

HS: Of those that do use an attribution model, the top 3 most commonly used were multi-touch, market mix modelling, and last touch. To what extent are these models helpful for marketers trying to measure the effectiveness of their social media campaigns?

GK: Facebook’s attribution and other last click methods are probably the wobbliest method for evaluation. Facebook’s claims every sale where someone viewed a facebook ad in the last 24 hours, or clicked on a facebook ad in the last 28 days, as driven by the ad.

For many brands its misleading to take these figures at face value. They don’t take into account other things that make people who view ads actually buy something. Covid 19 is a case in point. Brands that we’re working with have seen huge swings in their figures during the pandemic. Some seeing cost per acquisition go hugely up, and some hugely down. It’s clearly not because the platform has suddenly got better or worse at marketing, it’s just because the pandemic changed the pattern of demand.

What this means is that last click can absolutely be used to give a directional indication about whether this objective, placement, or platform was better than another. But because it doesn’t disentangle the role of social from other factors, it’s a bad idea to use it to decide how much to spend. You aren’t really getting the return on investment it reports.

HS: The brands you work with that really have their house in order, how are they approaching social media measurement?

GK: They do an intelligent combination of market mix modelling and attribution. Market mix modelling can disentangle the role of different drivers, not just social, but also the external environment e.g covid, price, deals, availability, and the competition. The most sophisticated marketers use this to set budgets for social, and then use attribution to portion that budget out across granular choices within social.

We also walk clients through a comparison between the outputs of market mix modelling and attribution because there are lessons to be learnt where their findings are different. The process often leads to clients making significant savings that can be redeployed for better sales results.

HS: And finally, what are your thoughts on the future for social?

GK: I don’t have a crystal ball, but my best prediction is that social has a really important role to play in future.

It’s primarily because of the ongoing decline in viewing of live TV and so of rich, audio-visual advertising – especially amongst young people. This trend isn’t going away and it has become more important this year because the pace of change accelerated during covid. 3 million people used Netflix, Amazon, or another subscription VOD service for the first time, and 96% of them intended to keep their new subscriptions (OFCOM figures).

Social, particularly YouTube is well placed to plug the gap. Social is generally more effective in categories with younger target audiences. And we’ve seen, across a number of clients, that it can act like TV. It can drive strong uplifts in sales, and often has a lasting effect on people’s behaviour.

Facebook and Instagram ads are currently less able to play this role. People can and do scroll past an ad so quickly that they spend only a fraction of a second with it. That makes it harder for them to truly generate demand, so that often, they only act as a reminder, or a signpost to help people find e-commerce sites.

My prediction is that facbeook and Instagram will be important after covid, because more shopping will take place online, so signposting to e-commerce sites will be more important than it was. Equally these ads will continue to support smaller firms.

In the longer term, and in absence of innovation that provides a richer ad viewing experience, I expect big advertisers to spend less on facebook and instagram than they do now, because they’ll refocus on ads that genuinely build demand.

Find out more about our work here, or read the next think piece here.