Question

Topic: Social Media

Need Ideas For A Social Media Experiment

Posted by michelle.woolfolk on 6/22/2010 at 1:51 AM ET 500 Points
My company is in the data & analytics business and we have an interest in doing some kind of social media type experiment that would allow us to demonstrate how new data sources from social media, web, mobile and traditional transaction systems can really change your view of your business ie.. customers, products, financial information, operations etc.. We have our big user conference coming up with about 3000 attendees that we could target to participate in an experiment.. but we are stuck on what that experiement and the theme of it should be. It needs to be interesting so people would want to participate in mulitple ways (surveys, social media tools) and include different types of information sources (our customer data base, customer registrations, survey's, social media tool participation, purchased data etc.. that we integrate to reveal interesting trends and the value of integrating all these data sources for a better view. Obviously we have privacy issues to address, so the results would be more aggregated trend related vs. specific person detail. Happy to hear your ideas .. feel free to ask questions.
To continue reading this question and the solution, sign up ... it's free!
SUBSCRIBE
Or, SIGN IN

RESPONSES

  • Posted on 6/22/2010 at 10:27 AM Accepted
    I think the way to approach this might be to deliver a consistent message to different people using various media channels -- including social media.

    The analysis would be to see how people respond when they get the message via one medium vs. another vs. multiple media. Response could be a simple attitude and awareness study, or a brand image audit.

    And you probably want the study design to be handled by a trained and professional market researcher to be sure that the variables you intend to test are, in fact, the ones you test. This is almost like a behavioral economics experiment. Not suitable for amateurs.

    If you want further thoughts, contact me offline using the email address in my profile. This is a very interesting question.
  • Posted by betterwords4you on 6/22/2010 at 10:35 AM Accepted
    No need to make this experiment very complicated or widgety.

    1. People get involved in social media because they want to feel connected and engaged with other people, and hopefully learn something in the process.

    2. Technology has changed how we engage with each other, but not why we feel that need (especially Americans, the world's most avid association junkies)

    3. The more AUTHENTIC the initial online engagement--the more relevance, proof, and value in first interactions--the more a person wants to stay in touch with another individual or group.

    4. So divide your participants into three or four random groups and have them read several posts, communications, messages etc., each of them written in a style similar to one of the major social media platforms (Facebook, Twitter, LinkedIn, YouTube etc).

    5. Then have all participants complete a survey with 20 or so questions, each with five multiple-choice answers. This will help determine how engaged respondents felt with the writer or presenter: how interested they might be in meeting that other individual in person, post a comment, refer it to a friend, etc.

    6. Since you're so clear on your objectives, it wouldn't be very hard to make sure your survey also gathers some data about the customer data base, customer registrations, social media tool participation, purchased data etc.

    7. Make sure you have each respondent indicate at the beginning how often they already use the various platforms in order of frequency, so their answers can be weighted accordingly.

    Good luck and let me know if I can help! (Google my name, Carey Giudici, to learn more about my background, read articles etc. if you're interested)
  • Posted by michelle.woolfolk on 6/22/2010 at 10:49 AM Author
    Thanks for your input, I think you have all outlined some interesting approaches, what I am still stuck on is the theme or maybe the motivating questions that would provide interesting results/story. For example, most of our targets for this are also our customers/partners and people that know our business and technology. I was thinking that maybe the experiment is focused on letting people submit their ideas if they knew X (had the data to support) they could do Y to change their business, their lives or their world. They could use social media and other tools to sell their idea to other participants and maybe there is a reward of some kind at the end. This isn't quite right, but its the theme/ purpose of the experiment that has me stuck. The more people are engaged the more information they share and the more analysis & insights we can capture. Hopefully this provides more perspective.
  • Posted by modza on 6/22/2010 at 11:23 AM Accepted
    I'm going to try to narrow down the range of possible ideas by making some assumptions. Of course you should correct these, but at least you should be able to follow my logic.

    I'll assume your list, "customers, products, financial information, operations" is a list of suggested areas of impact, and you don't really expect one experiment to demonstrate changes in one's view of all of them at once.

    While I believe that social media MAY have an all-encompassing impact on a company, every company's stage of understanding and reacting to the modern world is different, and you're not very likely to see all those impacts at once. Also, who is the person representing each client company? If they're relatively low-level, they're going to have a tough time selling the heads of each department on the changes needed, even if the experiment is an unambiguous success.

    I'm also going to assume you want to do an experiment for which the results make your case -- not quite a true exploratory experiment, but perhaps one you can run in test form first to confirm your hopes, then run on a large enough scale to be convincing.

    I'll assume that your existing customer base (very impressive size!) is not already convinced about social media, so the goal is to show them they need to add social media analytics to the tools they're already buying from you. If that's the case, there's a reasonable possibility that they are below-average participants in social media. Do you have any easy way to check? Look for a sample of names in social media sites?

    You proposed running the experiment on the (presumed) skeptics themselves and thereby making it doubly convincing, but it's more likely that your customers' customers are who they're most interested in. So how diverse are their markets? Are they all (or mostly) b-to-c, at least? Is one market segment viewed as a bellwether -- "if those guys are doing it, we should do it too?"

    Finally, or perhaps this should have been at the top-- what do YOU think social media actually reveal about any of the items you listed? The most typical use cases are: customer service (think @ComcastFrank); product development (crowdsourcing a la Starbucks, Dell, Zazzle); branding/image (Ford?); sales (Ford, for sure)... What else? Early warning system (get results before traditional surveys/panels); cheaper: than focus groups, for example... After all this, you probably are still thinking, okay, so give me an example of an experiment already! Well, I have to get back to my own work, so maybe later ;)
  • Posted by michelle.woolfolk on 6/22/2010 at 1:42 PM Author
    Thanks, you are right we are trying to figure out what we can reveal based on the data we know we can capture and what we would like to capture.. I have been trying to look at the big picture and what do we want to accomplish.. theme of experiment.. maybe we just need to look at the data we can get and the insights that we could gather from it that would be interesting.
  • Posted by modza on 6/22/2010 at 2:00 PM Member
    I appreciate very much that you want to tackle the biggest possible problem(s)...So IF your attendees are equally motivated (an unfortunately big "if"), then perhaps you could run two parallel branches, an "old-school" control group, matched for whatever you can (age, rank, industry, say, against a social media branch. Start at the beginning by asking what world issues are important to each group, allowing comments, picking the top one, then crowd-sourcing ideas for solving them, allowing comments, and finally ranking the ideas. Then measure how many people outside each group became aware/involved (social media reach), how many inside each group participated (response rate) and how far along the funnel of action they went (engagement, call to action) -- and perhaps how good the ranked answers were, by some external group of experts in the field...How fast you got through the process in each group, how much each cost (or would cost)...What do you think? Or perhaps set groups from one industry to solve problems within another industry -- if world issues are too lofty for your attendees...What do you think?
    Michael Odza
  • Posted on 6/22/2010 at 2:07 PM Moderator
    My suggestion: Start with the end in mind and work backward from there.

    If you know what you'd like your presentation to look like, and what you'd like the take-away to be, you can design the research to either support or refute the hypothesis. And either way it will be useful and interesting.

    Write the presentation you'd LOVE to make. Then let's see what it would take to get the information to support it ... or what compromises will be needed to deliver a useful result.
  • Posted by steven.alker on 6/22/2010 at 2:55 PM Accepted
    Is that your Teradata partners use group convention? If so you can’t afford to get this wrong as they have a) Paid to come and b) Are your lifeblood!

    M Goodman and a couple excepted, some of the answers look like the types of essay we used to make up where the aim was to do anything other than write about the subject and say nothing at all. We were always caught out!

    Unfortunately big ERP and Data warehousing produce huge expectations, ask you clients and I am sure that your partners will feel the same. 3000 is a hell of an audience to impress, so best not to cock up!

    Maybe you need to look at both the questions and the mode of delivery so that it has ongoing wow factor and holds attention a bit longer than a short sharp survey would, regardless of what you ask them.

    Take the example of Dr Richard Wiseman, a very clever lecturer in psychology at the University of Hertfordshire. He wanted funding for his conventional work but also wanted to get backing for his interest in ESP (Extra sensory perception) so he approached Twitter and some software manufacturers with a really off the wall proposition which worked.

    For his annual convention, he got 10,000 volunteers, via FaceBook and Twitter to join in a massive if unscientific ESP survey. After announcing the project 6 of his colleagues went away to spend a day in a particular environment and had their details published on the web for the volunteers to scrutinise. Then via Twitter, once they were in position the volunteers were told to focus on one of the six at a time and to try to work out what their environment was like by ESP. The volunteers responded via twitter and the experiment was refined on the hoof to ask the volunteers if they felt that they had ESP powers. As the results came in, in real time they were shown on a huge conference screen with a running commentary from others in his department and technical input from Twitter and the software supplier.

    It nearly brought the conference to a halt but if he’d been selling the technique he would have just won 3000 new customers and the next day the serious press ran in-depth stories whilst the tabloids made things up as usual.

    It is technically very challenging but for Teradata it should be a shoo-in! Try something off the wall like that – the more at right angles to reality it is then the better it will work. Or just borrow Dr Wiseman’s experiment – being an academic it is all in the public domain! In fact, invite him to repeat it for your convention but watch out – he’ll persuade you to fund him!
  • Posted by michelle.woolfolk on 6/22/2010 at 3:06 PM Author
    I think your thinking mirrors mine, but I don't think we'll go down the ESP path.. although I wish I had ESP.. but something relevant to shopping, fund raising, customer services that would resonate with our customers and could be applied to their business ..would get the most bang for the buck. yes it is our Teradata User Conference and yes it is a big deal.. so we have to get it right and show that when you integrate data from multiple sources about something.. you learn more about that something.. than you do if you only look at it from one channel or source. What is that "something" that is interesting and compelling.. Maybe I do need an ESP specialist.
  • Posted by michelle.woolfolk on 6/22/2010 at 3:07 PM Author
    I think your thinking mirrors mine, but I don't think we'll go down the ESP path.. although I wish I had ESP.. but something relevant to shopping, fund raising, customer services that would resonate with our customers and could be applied to their business ..would get the most bang for the buck. Yes it is our Teradata User Conference and yes it is a big deal.. so we have to get it right. We have to show that when you integrate data from multiple sources about something.. you learn more about that something.. than you do if you only look at it from one channel or source. What is that "something" that is interesting and compelling? Maybe I do need an ESP specialist.
  • Posted by modza on 6/22/2010 at 4:20 PM Member
    I go back to using social media -- and not just MarketingProfs -- to identify the best question, and follow all the way through. Ask your customers -- or a select sample-- what experiment they would like to see. That's the whole point, right? Not depending on us "experts" but engaging your customers/potential customers. (Ask them via social media, of course!)

    Then ask for help designing the experiment, conducting it, even reporting it.
  • Posted by modza on 6/22/2010 at 4:25 PM Member
    Charlene Li, co-author of Groundswell and now of "Open Leadership" is speaking now at BtoB Mag's Virtual Tradeshow. The key point: to take full advantage of social media, you have to know how to work without having full control!
  • Posted by michelle.woolfolk on 6/22/2010 at 5:51 PM Author
    Yes, I have read Charlene's book.. I believe she was with Forrester. Maybe I should look at it again. Thanks for the feedback.. will keep plugging with my team to work it out.
  • Posted by steven.alker on 6/23/2010 at 8:22 AM Member
    Conducting a live multi media survey of 3000 partners to assess their views on the future of data warehousing mining and forecasting would be about as interesting at a convention as watching paint dry – and I’m from the industry with more than 100 white papers to date. Believe me, if I wanted to I could bore for England on the subject but I choose not to!

    If you go back to marketing basics rather than this amorphous rubbish about aligning yourself with web channels (What with a spirit level?) then you start to break this down into ideas which pose questions which in turn have answers relevant to this project – by the way if someone had handed this to me when I was 29 I’d have either been the most excited marketer on the planet or Id resign – all depends on whether I thought I could do it or not!

    Beak the proposal down into its constituent parts:

    Exposition or going out to your audience – how and what to ask them, through what media? How are you going to ask your audience to participate? Email them? Text? Notice on your website@ The ideal would be having all 3000 take part because they want to and then discover to discover that participating was the most interesting and illuminating thing they did that day or that conference.

    Platform: Pen and paper and someone entering the data? Unlikely, but decide if the participants will respond through a laptop, a notebook or a mobile. Will they respond through the same media which you used to sign them up? If they do are there finesses you can demonstrate to making their data private whilst still collecting personal marketing information and crunching numbers.

    Data collection and manipulation – often a black box function but if you do this in real time, you can make the intermediary steps available in real time – better than just showing an accumulating graph which shows that your delegate’s truly believe that New Teradata does Wash Whiter, show them the engine room and some of the calculations and intermediate results on a suitable graphic

    Presentation and Analysis – again on a big screen and in real time, the last thing you want is something which shows that the answer is 42 for 3 hours. When our would be leaders were grilled in debate for the first time, some bright spark produced a voting panel of neutral voters and then measures how their opinions swung behind a given candidate as the debate went on and relayed it back through Twitter. The BBC went one step further and extrapolated the results into a model house of commons. Ask your interviewees some questions which indicate their views over time so that you can do trend analysis – you know, How important was this to you last year, now and in the future?

    If it had been the Open University they would have shown how critical processes were taking place in order to interpret the data and turn it into information.

    And that last one is the challenge. Any idiots can collect data. It takes skill to ask the right questions and analyse the results but it really marks you out from the crowd if you can turn the lot into relevant information. A key way of securing this step is to ask of each of the 4 stages outlined what is perhaps a significant feature of a Teradata product or service and then show though this example how you deliver it. 4 key messages all from the Teradata marketing plan culminating in a bit of razzmatazz and a conclusion should serve you well.

    OK. Now can I have the job!
  • Posted on 6/23/2010 at 12:09 PM Moderator
    I'd hire you, Steve, if for nothing more than some great entertainment! :)
BACK TO TOP

Post a Comment

Related Discussions

Community Info

Top 25 Experts
(Social Media)