Optimising Humans! The Art of Data-Driven Social Marketing #SESLON
Marty Weintraub is the founder of US digital marketing agency, Aimclear and he has presented at some of the biggest conferences in the world, including MozCon, SMX events around the world and PubCon. Marty also writes for Search Engine Watch and Search Engine Land and is the author of ‘Killer Facebook Ads: Master Cutting-edge Facebook Advertising Techniques’.
Marty started the session by giving away a copy of his new book, which is called The Complete Guide to Community Management.
Marty started his session talking about how social media has grown and developed in the last few years and referenced how people’s views on it have changed over the last few years. Marty also outlined how people will never change, channels will change and there will be trends, however people won’t change.
Marty also talked about using social media data to know your audience / demographic. Marty used Facebook’s social graph to provide data for a specific sector – outlining the interests of a set age group and outlining how it can help businesses.
If you’re using Facebook ads and you’re not segmenting users, you’re missing an opportunity to understand your social graph and your customers.
Most community managers don’t know very much about meeting people – Marty ran a survey recently and most community managers don’t do much outreach – they just speak to people who ‘drop by’. In real life you go to places where your demographic are, which is what community managers need to be doing.
If you’re confident about the content that your business outputs via social media, you can get anyone to friend / love you.
In order to make friends that are going to be useful for you business, you need data! Marty suggests looking at the content being liked and find out who’s liking what specifically. Look at the content that users are liking and have a relevant conversation with them using this knowledge / data.
Marty recently looked at what a user was saying to a US shoe retailer and contacted her thanking her for her contribution and saying it was helpful. He then asked her a question and he got a long response back. She ended up ‘friending’ him and sharing his content to her followers.
Marty applies a ‘know me love me’ rule to social media, which basically means if you know someone and are willing to ‘friend’ them, you will love them.
You need to map keywords to conversations and users – this is what Marty trains community managers to do. Everyone knows how to look at what keywords are converting or delivering quality traffic – we need to do the same within social media. You need to look at conversations and topics that are going to generate value.
One thing that all social media communities have in common is that they all restrict you in navigating social conversations, partly due to privacy and partly because it will prevent you from wanting to do Facebook ads. It’s a problem in international social media communities too, all of the networks have the same issue – finding real-time conversation is impossible. Having comments about specific topics that’s recent is extremely powerful as you can interact people and generate ‘friendships’.
People use Tweetdeck to track real-time conversation and respond to queries or brand mentions. Twitter advanced search is better – you can filter in a lot more depth and you can specific the types of content you want. You can also filter non-commercial content and find real users.
Hootsuit also has some great features – it allows you to do negative filtering. Hootsuite is the most advanced for filtering content and it’s really good for sentiment analysis.
Marty was really excited about graph search for finding conversations – however it’s limited by degrees of separation. One of the main issues with graph search is that you can filter by recency.
“Site:socialnetworkdomainname KEYWORD” is the best resource for finding themed and relevant content to interact with – it’s the best tool around for the community manager. This operator can be used within all major search engines, not just Google.
Google+ is really bad for searching – it’s really hard to find the right people and conversations to interact with. The best way to find the right people / conversations in Google+ is to use the site search operator – you can also use Google Translate to make it more effective. Discovering how people are talking about your products in foreign search engines is really useful data / information.
Social distribution is not free anymore – if someone interacted with a brand on Facebook in the past, their friends would be able to see it. Now, if you interact with a brand, your friends aren’t going to see it. Facebook said that they did it because too much stuff was being posted, whereas Marty thinks there were other reasons behind it, which come down to monetization.
Paid organic amplification – this is very interesting, most of the social networks have paid sections of their website that they’re making look like organic content. They don’t want people to understand how much of a paid ‘cesspool’ the networks have become.
It’s important to create posts tactically to ensure that the content looks right if you run page post ads. Ironically, serving page post ads to your own fans costs more than it does to serve them to everyone. Running paid content ads through your wall will cost the same as a generic brand advert, but you will get more relevant and valuable followers.
Marty ran a targeted content ad on Facebook and it generated coverage on AdAge and lots of other places, it cost him $18 – all he did was marketed the content to the right people.
Phychographic research is to social media what keyword research is to SEO. Once you’ve got information about a user, you can get much better results from really targeting them. Example: selling late night pizza delivery to people who drink and smoke weed gets a really, really good conversion rate.
It’s important to find associations within demographic research – Marty filtered a group of people by occupation, where they graduated from and some of their product interests and targeted them with high-end music tickets, which worked.
Targeting similar budget products works well – Marty provided an example of targeting Rolex fans with Porsche ads. Occupation targeting is also a great hack, specificing precise interest alongside side company allows you to reach people by their occupation, so you can targets managers at IBM for example. You could also have a subscription that tells you their occupation (eg: if the user subscribes to physician weekly).
Look as deep as you can when targeting people, look at their interess, jobs, friends, likes etc.
Social is just like SEO, rankings are important – connecting with very social and influential people helps to increase your reach, these people are important. Marty recommends tracking your reach based on friends and second-degree friends.
Share of voice is a lot more than just volume, it’s important to really dissect this to understand who your key influencers are.
Top takeaways from Marty’s talk:
- Targeting is a risk
- Psychograpics, more than Facebook adverts
- Reach out – take your time and do it properly
- Follow public data and engage with people
- Be charming with people
- Be creative in finding the data you need
- Chrome translations are your friend
- Amplify content to psychographics
- Journalists use social media too
- Look at occupations
- Measure who’s in your community – don’t just look at frequency, share of voice is hugely important
- Don’t be creepy – it’s really easy to do it as a marketer