Like Minds Panel Discussion: Trust and Social Influence, featuring Azeem Azhar



Joanne JacobsJoanne: There’s a massive difference between influencers and celebrities. You shouldn’t try to make influencers behave like celebrities.

Azeem: We all know what a celebrity is. What’s an influencer? Lots of people are influencers – a large percentage of the population depending on context.

Davide: Advocates are the people who sit between celebrities and contextual influencers. Red Bull does an excellent job of locating these advocates. They are near you, they enjoy your brand, they want to be part of it. Companies need to understand this.

Azeem: Social broadcasters – Robert Scoble, Lady Gaga – lots of reach, not much trust. ”Mass influencer” – the next step down. Not as much reach, but way more trust. Prior to social media it was hard to do the large scale indexing and analysis that allows you to plan a campaign like this.

Chris: A lot of this is from the marketing point of view, The problem of trust is from the consumer point of view. Can you trust the motivation of people recommending a product? Do they just want you to think they own a Porsche?

Benjamin EllsBen: Does trust capitulate under inspection? If we find that everyone’s an influencer, we’re back to traditional marketing.

Azeem: It’s about context.

Ben: Celebrities are bizarrely influential. As much as family, some research shows. I shouldn’t be influenced about business software by a sportsperson, but people are.

Joanne: Influencers don’t like being treated like celebrities – they don’t like being told what to do. They hate product managers telling them what to say on Twitter. That’s why you shouldn’t treat them like celebrities.

Ben: We’re good at spotting people’s motivations – and we discount influence when people are being paid.

Davide: Suggests that influence can be subtle –  studies seem to show that people don’t recognise how much they respond to subtle signal of influence, including those from celebrities.

Azeem: You can now run experiments using the data and analysis you can draw from networks. People don’t respond to rewards? That’s not what the data shows. They do Tweet about things they are given. It’s worth testing this stuff out and seeing how it works against the objectives you set.

Chris: There’s a difference between rewards and payment.

Ben: The most powerful driver of influence is reciprocity – giving somebody something without asking for something in return. Influence is very different in different cultures. Not just different countries – different social groups. Brands see things as product categories, people don’t.

Joanne: Celebrities can give you a spike of awareness (journalists the same). But they’re not interested once it’s no longer new. Influencers keep the conversation going over years. Trust is built over a long period of time. Influencers are not part of a campaign, they should be a long-term relationship.

Chris ArnoldChris: My experience is that people are much more vocal when they have negative experiences. This is where we get into the sentiment area. That’s going to be a big area over the next few years.

Azeem: When you start to think about trust and influence, things get complicated. Someone who is influential on cardiac drugs is going to have positive and negative views. If you’re a constant negative detractor, people stop paying attention to you. You’re like a loon shouting on the corner. You’re looking at long term behaviours to make a judgement on trust.

Ben: Linguistic concepts like positive and negative are really difficult. From a linguistic point of view, it’s hard to tell if it’s a situation thing, or something to do with the person themselves. Is “expensive” a positive or negative attribute? It depends.

Joanne: Negative feedback is not always a negative thing. It’s the cheapest market research you can ever get. If you only produce positive attributes around a brand, people stop trusting you.

Azeem AzharAzeem: If you’re in sales, you’re only interested in messages that shift product. If you’re in product development, you’re interested in different messages. You might be more interested in negative messages.

Ben: People think of influence as a one step move – find influencers. Done. It’s a multi-move game. Turning around a negative influencer can be very powerful.

Davide: Social media is not a mathematical system you can predict. You have to work in cycles: try something, analyse it, figure it out why it worked (or didn’t).

Joanne: Doesn’t like Pinterest. Thinks it will have a limited life. Maybe it will have a role in experiential capture, bringing together images from an event.

Azeem: You’re be insane to go to a small business owner and say “ignore the 25% growth from Pinterest, it’s not long term”. They should experiment fast, use it if it works, and get out fast when it stops working.

Ben: Different people like to be communicated to in different ways. In some businesses, the spreadsheet is the religious object than makes things happen. In other businesses it’s a story. There’s no point in communicating in a way they don’t want to be communicated to. There’s only one way to appear genuine: be genuine. The reality is what worked yesterday might not work tomorrow. Infographics have gone from nothing to over-exposed. But it’s become a massive “need” that companies think they need to have.

Azeem: There’s a difference in measurement requirements in different parts of the market. In PR you’re running a small number of relationships for your clients. In that situation, you’re best to go in and run that manually for your clients. Just focusing on a single figure is not likely to be useful. Figure out what you’re trying to achieve, and then how you need to measure influence for that. PeerIndex is trying to optimise for sample sizes of 100+, not for individuals.

Joanne: All these tools are based on your behaviour, not on what you are an expert in. The things you share online might be what you’re interested in, rather than what you’re expert in. And that will chance how the influence tools perceive you.

Ben: The difference between what people click on against what they share is a very interesting measure…

Chris: Influence metrics are not accurate yet – but they’re good enough that you’ll get RoI from it. You’ll miss people you should have talked to and talk to people you shouldn’t have. But you have to test what you’re doing, and other approaches.

Lots of debate about the validity of measurement. Ben points out that you can call things into existence just by measuring them. You can create things that don’t really exist in the real world. Joanne cares most about her immigration score – and so cares about her credit score (which half the audience claimed not to – to some skepticism). She has her doubts about these measurement tools – but she is glad that they exist, so we can start to explore how people influence each other.

Davide CasaliQuestion from the audience: Should influence measures be used for recruitment? In some places lack of influence is a real problem.

Azeem: We do strongly prefer people with a good Gitalytics score in the areas we work.

Ben: Numbers are part of a story – understand what that story is. And remember that people like to be valued, not measured.

Davide: These tools are evolving. As you add more information, ratings change and shift around. No-one knows where this will end up.

Originally published 17 February 2012 on Like Minds