Hi Dennis - thank you so much for your kind words. The common calls for determining ROI is (IMHO) largely about helping people who don't "get" social media determine value. If you can show numbers that justify the expenditure to an exec who doesn't understand the implicit value in these type of relationship-building activities, you'll get the green light. After all, what's the ROI on lunch? We don't feel the need to justify that, since we understand well the cost vs. value of personally connecting with clients and other people who are important to our businesses.
That being said, for a company of SAP's size, even showing a minute increase in some aspect of revenue generation would more than justify an extensive social media strategy.
Dennis Howlett
· 2 years ago
You're right Maggie and I've previously looked at this with my accountant's hat on when Charlene Li was doing her study (I contrib'd).
When I think about ROI, my problem is one of attribution, certainty and confidence. Can 'we' be certain what we're measuring, that it's directly attributable and the impact it has? I don't think so at this time. That makes me hesitant.
I would love to build a model capable of measuring these things but each time I look at the constituent parts I come up against brick walls.
That leaves me with the issue of credibility. Sure - I can build a model but if it can be busted then I've got a problem.
Charlene's final model was 'ok' but capable of demolition. I've not seen an update that makes sense to me as a cheque writing CFO type but hey - happy to hear if there's one around.
maggie fox
· 2 years ago
Dennis, I think Charlene's model is an excellent place to start, but we have found that calculating ROI tends to be "after the fact", i.e. several months post-launch of a program, evaluating specific incidents and (attempting) to calculate the ROI on the outcome. Of course, selling that is somewhat difficult, but programs can always be (strategically) pulled if they're not performing, an initial outlays tend to be fairly minimal.
Everything about this space is (IME) bespoke - there is no one size fits all (which is very like human nature, isn't it?)
Dennis Howlett
· 2 years ago
That's why you need us accountant types -:)
You can't logically measure ROI until after the fact. But what you can do is model in advance and then do continuous measuring exercises. None of that's possible without having access to the underlying data and in my experience, most companies are pretty clueless on that point.
I'd also add that marketers are the worst people in the world to attempt measurement (by and large) and IME. Another reason for bringing in the bean counters. Which was why I added my 2 penn'orth to Charlene's study on the topic.
I know we get a lot of bad press but most of us want to help...
All Comments
That being said, for a company of SAP's size, even showing a minute increase in some aspect of revenue generation would more than justify an extensive social media strategy.
When I think about ROI, my problem is one of attribution, certainty and confidence. Can 'we' be certain what we're measuring, that it's directly attributable and the impact it has? I don't think so at this time. That makes me hesitant.
I would love to build a model capable of measuring these things but each time I look at the constituent parts I come up against brick walls.
That leaves me with the issue of credibility. Sure - I can build a model but if it can be busted then I've got a problem.
Charlene's final model was 'ok' but capable of demolition. I've not seen an update that makes sense to me as a cheque writing CFO type but hey - happy to hear if there's one around.
Everything about this space is (IME) bespoke - there is no one size fits all (which is very like human nature, isn't it?)
You can't logically measure ROI until after the fact. But what you can do is model in advance and then do continuous measuring exercises. None of that's possible without having access to the underlying data and in my experience, most companies are pretty clueless on that point.
I'd also add that marketers are the worst people in the world to attempt measurement (by and large) and IME. Another reason for bringing in the bean counters. Which was why I added my 2 penn'orth to Charlene's study on the topic.
I know we get a lot of bad press but most of us want to help...