Why Social Customer Service Needs a Dose of Reality (Part 2)

This is part two of a post from Martin Hill-Wilson of Brainfood Consulting. Part one can be read here. Martin is the trainer for our unique 1-day training...

This is part two of a post from Martin Hill-Wilson of Brainfood Consulting. Part one can be read here. Martin is the trainer for our unique 1-day training course on ‘Social Customer Service’.

In part one of this post I explored the performance and perceptions of customer service, comparing social media to voice and other channels. I left off on the point that, perhaps, social customer service is really about quality and culture, rather than sheer responsiveness.

Call centres have a justified reputation for being obsessed with their own agendas over those of the customer.  To be fair, some have broken free, however many remain stuck. Meanwhile some brands doing Twitter are brilliant examples of being human, approachable and authentic.

  • O2 during their network meltdown in 2012
  • Virgin Money’s CEO who promotes herself as part of the service team
  • The oddball collection of advisor photos on the Ford Twitter site which act as a perfect foil to ‘anonymous global brand’ syndrome

When we are looking at this category of examples, social customer service is breaking new ground. No doubt about it and one which the incumbent customer service industry should learn from. Where I do disagree is with the current assumption that social channels are clearly and always better than call centre delivered service experience.

And I will tell you why I think this.

Going back to my opening point, I do a lot of reading. I’m always pleased to find examples of organisations being praised for being ‘good’ at social. Often this is described in terms of a key metric that appears on everyone social media dashboard. It’s that wonderful sounding word – Engagement.

Engagement is an interesting concept as used in the social media space. It indicates we have the customer’s attention. They have responded to us. The like us, they re-tweet us etc.  Remember we are still in the phase of Marketing rediscovering how to be relevant to an empowered consumer who is low on brand trust.

While these metrics are easy wins, in truth these are pretty superficial indicators and do not match the original meaning of the word engagement which is more to do with the feeling of being involved in something that matters.

The reason I mention this is that the examples I keep stumbling across of great social customer service are being praised for great ‘engagement’. For a while I simply accepted this and used them in my seminars and keynotes. However one day I decided to go look for some screenshots to show what this engagement actually looked like since we all learn better with real examples.

I was not impressed.

I’m not going to name and shame individual brands since it is pretty easy to go find examples for yourself and take away the right lessons. However they are some principles worth mentioning.


Social customer service loses a major advantage when advisors are allowed to simply cut and paste their initial response. That was what we criticised call centres for. This time the impact is even worse since everyone else can easily assess the house style and so this formulaic approach looks even worse.

Even if the call to action 99% of the time is to suggest that the customer direct messages you (which to my mind is done far too often) we need to make the effort to remain fresh and original for each individual customer.

And as a final comment on this topic, there is one well known brand that caveats every interaction with the word ‘appears’ as in “I’m sorry that it appears….”. No doubt this is the influence of Legal and their wish to impose up front damage control. However I noticed one exchange in particular in which a customer simply lost it around being patronised etc. So beware.

Continuous Improvement

Most customer service is about solving the same old problems. That’s why Interaction Analytics is now making such headway in traditional call centres so we can ‘stop doing dumb things to customers’. The same applies to social channels. In fact even more so since the repetition of doing ‘dumb things’ is clear for all to see. The impact is to make the brand look unresponsive and uncaring. Brand managers take note please and intervene.

Let me give you an example. Twelve out of the top twenty UK retailers are on Facebook. Where ever a point of sale exists, the need to provide service quickly follows. My impression from sampling a few of them is that delivery issues are the most common and contentious issue for their Facebook customers.

Omni-channel retailing is logistically tough and we all know retailers are fighting to survive. It seems many have tried to economise on delivery costs. However these problems then turn up en masse online for all to see. They need to be fixed. Certainly in the context that all retailers are currently fighting like cats to deliver the best customer experience.

Where is the feedback loop from the social customer service team back into the business? Who in the customer experience team is sweating to make fundamental improvements yet meanwhile is sorting out short term compensation tactics?

This is still a tough act of co-ordination within mainstream Customer Service. Let’s hope socially sourced customer insight receives better treatment.

Last Thoughts

If you have managed to read this far, thanks for staying the course.  I hope I have offered some constructive criticism for us all to consider. Each year the optimists amongst us proclaim this is the time social customer service goes mainstream. Personally I would hope we fix a few things before more customers get exposed so that the vision for social customer service is realised before it is completely over egged.

Martin Hill-Wilson is running a unique 1-day training course on ‘Social Customer Service’ in London on (12th March 2013). Places are limited and tickets are available here.

In this article

Join the Conversation


  1. Jerome Pineau Reply

    Hi Martin — another compelling social service post. Thank you! On the engagement side, I do think it’s about being conversational but truth be told, on a service channel, people aren’t usually inclined to chat ad infinitum. Typically the last thing they want is to chat and hang out — they want their problems solved in real time. I think measuring and driving towards pro-active service is more interesting — both emotionally for customers, and conversationally. Another thing we measure is “gratitude index” — the percentage of engagements where customers express some form of gratitude — this to me is a top indicator of how successful we are in a given channel.
    And finally, on the feedback loop thing, couldn’t agree more. Unless part and parcel of a social strategy, you’re only halfway there if you don’t use social to evolve and mutate your processes and rules to improve customer experience. This is where many large orgs fail in my experience. They think just doing social is enough to keep up with the Joneses. But fail to realize that being social entails restructure and internal DNA re-engineering 🙂

  2. Martin Hill-Wilson Reply


    Thanks for being first into this leg of the conversation. In response to the engagement issue, I’m firmly in the customer effort gang in terms of their ethos – make it easy, fast and effective for the customer. It is within that assumption that I’m pitching authenticity. Actually since it is getting too much airplay let’s find another word. What about ‘being real’ – i.e. attentive and polite. That will do me!

    Anyway, more interesting is your gratitude index. It occurs to me that since this is an unsolicited response it should be a better indicator than a more routine NPS type evaluation. Would like to hear your thoughts on that.

    Also if the numbers are not sensitive is there any way you can share your current gratitude run rate? If not, then any insight into how often it can be expected (assuming a customer centric brand such as yours), what you use it for, can you attribute it back to an individual for their performance reviews etc etc.

    On the final point, I’m delighted to hear you are off the starting line as far as learning and improving. How do you pull all your insights (ie every channel) together and action them? This is still a mystery to most, so a glimpse of how its done would be a treat.

    1. Jerome Pineau Reply

      Hey Martin, I think it’s really hard to tie social service satisfaction to NPS alone. For one thing, I’m not sure surveying on public channels is statistically solid. Doing so in a service/support community with well-identified registered accounts at the point/time of transaction makes perfect sense. But on Twitter/Facebook? Not convinced. For another, I don’t think people recommend a brand’s service arm per se — rather, they recommend a brand as a whole (service being one aspect of it) – So it makes more sense to me to apply NPS to the brand as a whole. But for sentiment on real-time streamed social channels, I intuitively feel NLP is a better approach.

      We examined our 2012 engagement stream on TW and did 20% GI rate on there. I blogged about the results (and how we look at social service metrics) at http://www.socialmediayousay.com/post/40669540149/really-interesting-article-from-eric-pratum

      The other cool thing about it is you can mine the same metric from other accounts and then compare yourself 🙂

      We have not tied GI to specific crew members or geos thus far but it’s not a bad idea actually.

      As for the actioning part, it depends on the KPI. If our 1:1 metric (posts tagged as deflected) wavers, then we know we have to improve response quality — or maybe send folks to more compelling support content. We operate based on “confirmed hit” worst case scenario for deflections. Meaning they are “absolutely” known — and we say ok, this is the minimum, we can assume our actual rate is double or triple (because most people don’t take the time to come back and thank us or tell us whether we helped them or not — such is the nature of help seekers ) – But to the execs, I can *prove* deflection with confirmed hits. On the TTA/TTR KPI (engagement velocity, resolution speed) that’s easy because those modulate our performance requirements. If we slip there, it’s easy to find the reason and adjust process and/or human resources adequately. Using bit.ly to track content and reach/clicks helps us adjust our publishing “machine” on the content side and optimize our P2P community routing tactics.

      Initially we were just looking at ops metrics — but those are useful to setup and run the business as a traditional support shop, not necessarily a social service one. Volumes, sentiment, topicality — few people except “field commanders” ended up benefitting so then we aligned our social KPIs with the overall divisional targets — then it’s a matter of mapping the “right” social numbers to each selected business goal/KPI. Of course added benefit is that these speak to the execs 🙂

      Sorry this was probably too long an answer – what happens when you get me started on social service and metrics 😉

      1. Martin Hill-Wilson Reply


        Have now checked out your GI scores & KPI gameplan. Genius!

        By the way as an advocate of speech analytics, I think I’m going to introduce a GI index for verbal interactions.

        On a more general point, while Csat/NPS/Customer Effort are often corporately imposed and a given, in fact unsolicited GI is much closer to reflecting ‘value delivered’.

        What really strikes me though is the clear difference between the p2p community centred social service model with strong KM & Facebook/Twitter slotted into that context (often mature in tech heavy sectors) versus the rookie levels I was complaining about in my post (from retail). Thanks for reminding me that there is another end of the competency spectrum.

        On the actioning part you describe, that’s very interesting in itself. However I was referring to continuous improvement as opposed to operational BAU mgt. Your 1:! metric points in the direction i’m alluding to. Put another way, what can be taken upstream and fixed before it becomes a trad/social care issue downstream?

        It strikes me that dumb stuff looks even dumber if it’s still there month after month for all to witness (inc any competitors smart enough to be looking for weaknesses to exploit- must be a new outsource business in there somewhere-)

        Anyway do you spot the prob, then socialise it directly and thereby increase your network of social service ‘contact points’? If so, is there any attempt yet to compare and contrast across channels (trad & social) where issues show up and which need to root cause fixed?

        1. Jerome Pineau Reply

          So, on the feedback loop issue — continous improvement — I feel these kind of have to be different for business vs. tech support. On the business side, you have high-friction processes, rules, regulations and “we’ve always done it like this” types of pushback. Social is very good at surfacing these but so is common sense and basic customer advocacy 🙂
          And in the end, nixing these is a political and sometimes business model issue. And, if you cannot or will not fix these, then at least communicate as to WHY — so people don’t feel they are arbitrarily mistreated. Behind every stupid business rule and process is usually a historical reason. Customers can be forgiving if you’re upfront and clear and honest. Social is one way to achieve this type of relationship with the base of course.

          But social is only the messenger there. On the tech support side, it’s a different ballgame because you have crowdsourcing and ideation feedback loops and those can (must!) indeed be exploited via social of course — caveat there is that the product managers have to become “social” — there again at times you may have political pushback.

          On the “contact points” we tend to seek them out judiciously more than trying to be everything to everyone everywhere — the same issues show up in traditional and social venue – of course on social venues they are greatly amplified from a brand perception angle.


          1. Martin Hill-Wilson


  3. Why Social Customer Service Needs a Dose of Reality (Part 2) | Our Social Times | digitalnews2000 Reply

    […] on oursocialtimes.com Share this:TwitterFacebookLike this:Like Loading… This entry was posted in Uncategorized by […]

  4. Robert Bacal Reply

    Unfortunately, companies that use social media for customer service are simply taking their philosophy from existing channels, and making the same mistakes in social as they have elsewhere, which is why I’ve said (and I believe you’ll agree, Martin), that first address WHY your existing customer service channels are poor, look to yourself, BEFORE using social. We have more and more companies jumping on social and being as bad there as everywhere else. Here’s a quote from: Customer Service and Social Media – Myths and Myth Steps – http://work911.com/articles/custsocialmyths.htm
    “For large companies, attempts to use social media to address their poor customer service amounts to trying to put a band aid on a shotgun wound. It’s happening for a number of reasons, and the reasons have nothing to do with the tools used to deliver customer service, which is why social media doesn’t change anything at all at the business results level.”

    1. Martin Hill-Wilson Reply

      Sorry to have missed your contribution. Have just returned and are now chomping through your articles on the topic. I’m in synch on the core insights. Moving furniture changes nothing. The new zealots lack historical context for their enthusiasm and therefore are doomed to disappointment.

      I do see some large brands spending large because this new digital behaviour puts them on the back foot and they can see bottom line shrinkage.

      All of this though is about the impact that real leadership has on a brand momentum. In that situation even customer service pays for itself.

      Finally interaction analytics provides the ‘kill shot’.