Dec 20, 2009

Problems with Social Media Technologies

by Mark Rogers on 9 December, 2009 - 22:32

In the summer Hyper Happen’s Asi Sharabi wrote an anguished postabout dealing with the output from social media dashboards. Luke Brynley-Jones, who is putting together a panel at Monitoring Social Medianext month, asked me to talk about it. At the risk of “scooping” my contribution, this is what I wrote:
My sense is that there are several issues at work here. There are inflated expectations of what the software can do on its own. There is a general reluctance to appreciate the scale of the undertaking. Put at its most simple, each partner client, agency and measurement specialist needs to field folks with real expertise in:
- The brand and its online community
- Online engagement as a domain
- The chosen tool and its most useful metrics
- The client’s business, its capabilities for response
and to devote (as Asi complains) real time to it. It is messy, because no one tool performs all the tasks you need, and much of it requires painstaking manual cross-referencing. We never devote less than a day to compiling a monthly report of most of the brands we cover (often a day and a half or even two). 2-3 people work on it. We don’t just use our LiveBuzz monitoring tool, we use Google, Yahoo!,, search Twitter direct, in short cross-reference and check everything. Without a contextual report the value of any tool alone is hugely reduced.
Outstanding issues:
- Missing data (getting better, but always an issue)
- Flaws in identification and automation of topics
- Automated sentimenting that works for some contexts (high volume, generic) but is weak in others (low volume, specific)
- Conversations that are a propos but don’t appear in the tool’s on-topic conversation bucket (e.g. posts on a brand page, Facebook groups, petitions, backlinks)
On a higher level I think that Asi’s frustrations result in part from the job he is being asked to do. It is unsolvable by him. He is not being given the resources (time) needed to really add value here by addressing the big issues:
- The context: what can you benchmark this conversation against? What is the history of the debate? Is that data even available?
- The business value: how do you derive really valuable actions from the data? Who has the expertise to do it? (It ought to be the agency, but they are often too close to the fray, and have an agenda which can lead them to recommend marketing activities of some kind – and this may not always be the best response)
- Having recommended specific actions, who undertakes them on the client side? Is this a marketing issue, or a business process issue, or a customer service issue, or a strategic issue for the business? All those siloes of the business have to be bought in for the thing to work.
In short: social media monitoring tools stand at the bleeding edge of the social media “question” for the industry, a question to which no one has yet formulated the full answer.
That question can be summarised thus:
- How does a business take full account of the value of the insights that are now available to me from the net?
- What is the role of my existing activities? Are they all valuable? Which ones do I keep and which discard?
- Who takes those decisions within the business? How are the decisions validated?
- What become of all of the existing stakeholders: internal teams like PR, marketing, brand, investor relations, customer service? Can I reorganise around this change?
- What becomes of all my “talking” activities, via all the different agency partners, promotions, DM, above the line, below the line? Do I centralise them, consolidate them? Drive everything via an “insights and measurement” strategy?
My cheeky response is that Asi is a square peg in a round hole. The client is apparently asking him to become a Market Sentinel analyst without providing him with a budget! He needs an analyst to work for him to do the grunt work, so that he can add his value by addressing the big client-focussed questions.