For Questions on Twitter in Customer Service, Press 1

IVR systems get no respect.

An IVR could be considered a great addition to a call center.  It handles all incoming calls, resolves the simple requests for service through interactive applications, routes the calls to the most appropriate agent, and captures identifying information to pass along .  Is the perfect attendant and can scale to hundreds of simultaneous incoming calls.

It should be considered a success story.  Users don’t want to use it.

When first introduced – only functioning as a phone-tree with routing functions – people liked the novelty of it.  The novelty wore off and the system showed its true colors: an automated routing mechanism with little forethought put into at deployment time, no integration or interactive abilities, and few tales of success.

Vendors quickly began to improve their offerings, create better programming interfaces, provide more interactive functions and add voice recognition.  Today is far better than it was initially, and is actually useful to automatically solve around 30% of the calls.

Twitter is moving along the same path.

I referenced in the past the problems with Twitter and how you should only consider it another channel for customer service.  Research showed companies used it mostly as an escalation tool to the call center, or  to create tickets; few cases are solved via Twitter.  It  is closely related to the IVR. Focus on what Twitter can do well and stop thinking that it can do more than it can.

We will see innovations for the use of Twitter in Customer Service that will make it a better tool.  For now, use it as you use your IVR: automating large volume of interactions, routing and creating tickets, and provide a listening ear for feedback.

Anything else, simply cannot be done now. Do you agree?

32 Replies to “For Questions on Twitter in Customer Service, Press 1”

  1. I agree for now.. But there is more:

    Twitter can be seen and compared with an IVR and maybe even managed like an IVR. But there are more differences than similarities

    The first difference is that Twitter does not feel like an intelligent system. Its success is based on the fact that it feels like human-to-human interaction and engagement. And that’s actually what it is.

    Second, it requires less effort from the customer. Customers can just tweet the name of the company and immediately the companies ears are tuned to the person. The ball is now in the companies court. With an IVR it is still the Customer who has to make the choices, with tweets its the company that has to make the choices (do I engage or not? do I redirect to other channels? etc).

    Thirdly, as you point out: all the great intelligence built in IVR’s these days are not (yet) available for Twitter. This makes routing to the right skills (in high volumes) a great challenge for Twitterservice.

    Last, maybe most important difference: the main drivers (for companies) to use IVR’s is to avoid human-call handling for easy questions and requests (costs) as well as to direct customers to people with the right skill-set to deal with the customer request. The main drivers to use Twitter are improvement of the customer (brand) experience (by showing you are listening and engaging when necessary) as well as collecting WOM and Feedback.

    With so many differences I’m hesitating to say Twitter and the IVR are closely related. And if they are not, in their essence, related, what does this mean for future innovations?

    Will we ever have the intelligence in tweet-routing as we have it in the IVR? Will we ever be able to automate high-volume tweet-answering. And if we do, will we loose the personal engagement touch there currently is in Twitterservice? Will that not mean the end of Twitter as a Customer Service channel?

    My conclusion (for now):

    I think Twitterservice in relative low volumes (as I think it is now with most companies active in this area) stands a good chance to do the tasks you described (and not more).

    When Twitterservice becomes a high volume channel, it will become non-personal through automated responses or re-directions to other channels, with no added value over other channels.

    If it does not become that, it will be only for those companies that allow themselves to spend a fortune on Twitterservice, because it will be very labour-intensive.

    My question to you: Would it not be best, If a business runs a high volume service center right now, and they do not want to ruin the customer (brand)experience on Twitter in the future (when the masses discover the twexperience on low volumes), to just stay out of the Twitterservice at all, and only use it as a listening ear for feedback?


  2. Great comments Wim. I tend to agree as I view Twitter as more of a megaphone to announce your problems to the world whereas the IVR is simply a 2 way walkie talkie with only one person listening on the other end (public vs private).

    The problem that many people face is they see new tools as merely extensions of past tools and often try to use them the same. Twitter, and tools like it, are much different than IVRs and should be leveraged in much different ways.

    – Twitter enables mass problem solving, not limited to one company to one user. Leverage it by building a trained and passionate community of expert followers.
    – Twitter responses should absolutely be automated, to a point. If someone complains about voice quality on their VOIP system then an automated response should be tweeted to the user with a link to a simple trouble shooting page. The conversation should flow to the best tool to solve the problem which will often mean leaving Twitter altogether.

    I am curious to see what others think as well, great topic.



  3. Esteban,

    Good post and discussion.

    From the customer’s perspective, there seem to be very few positive IVR experiences. It is mostly an inefficient source of frustration.

    TwitterService carries with it the opportunity at this point to interact with a person with very little time or effort involved.

    To Wim’s point, this is probably a novelty at this point – but more and more people are looking to Twitter to be supported.

    Some companies are doing a great job of listening and engaging. Most still aren’t in the game, or to Esteban’s point are beginning to duplicate their mistakes all over.

    Best regards,


  4. We’re thinking along the same lines. I just wrote about twitter as a source for text mining “customer experience” data, and while it’s a useful signal for spiking customer service issues when analyzed, the low volume, lack of representative sampling, lack of inherent customer data in the tweet, and requirement for direct customer/company interaction makes it less transformational than all the hype might suggest, and more just an incremental way to do support.

    I believe that companies should focus on more direct customer feedback for intelligence/analysis purposes.

    If they can have a conversation with a tweeter, and identify the customer’s specific issues, provide diagnosis, resolution, etc via twitter, fine, but from a customer experience analytics perspective, we find at Clarabridge that companies gain far more insight and value” from call centers, survey feedback, and more interactive social media outlets such as forums, newsgroups, and review sites.


  5. Excellent conversation here! 🙂 I have learnt so much & got so many perspectives here! Having faced an IVR only as a customer I do not have rave feedback on them. 😉

    Though as a developer who dabbled into their basics way back in 2001-2 I was enthused with the opportunities to build so many useful stuff that would alleviate my pains as a customer, the technology itself did not deliver. I unfortunately got into BPM & couldn’t get into IVR development. 😦

    Have any of you looked at any IRC based support channels? Admitted its used by geeks, but its a model that can be modeled on Twitter too! They have bots answering FAQs as well as developers & experts (community of power users) answering other queries or routing to resources on the web.

    Look at Ubuntu IRC channels, it will give you a very good idea of what am talking about.


  6. Forgot to mention one caveat with IRC though … I guess Canonical guys did not consider the common man who would not know how to use IRC … so while their IRC channel has great model for customer service, its reach is limited. They are not on twitter, though has a !ubuntu room where the community responds back to you. But again is a geek’s arena. Twitter though more popular than doubt if it too represents the most popular channel of the target market segment.


  7. As a marketer, I absolutely love the idea of Twitter for customer service. I wish all my COMPETITORS would use it. That way I can find out what is wrong with their products, I can see how they respond to very public cries for help, I can spot likely high-value customers who are at risk of defection and I have a tailor-made ambush marketing channel to capture them too.

    Thank you Twitter.

    Graham Hill
    Customer-centric Innovator


  8. I think that the comparison of Twitter to IVR you are making is incomplete. True, if you think of Twitter as a tool or medium for facilitating conversation between two parties, the company and the customer, I agree that it will quickly run into the same issues as other service options and require automation inserted between the customer and company in order to make it scalable and cost-effective.

    However, that tends to miss the real value of social services like Twitter: it’s not about providing another way to connect the company and the customer, its about connecting the customer to other customers who can respond on the company’s behalf. Why force the customer to talk to a robot (no matter how well programmed) when you can help them talk to a volunteer expert who can provide more solutions than even the most sophisticated self-help systems, and the emotional satisfaction that a computer program never could. That a company can identify power users and advocates, and facilitate their interactions with the company’s other customers through a channel like Twitter is the real promise social tools have. With a community of volunteers you can achieve economies of scale that previously could only be accomplished through automation, but without any loss in perceived value by the customer – in fact, they often value the help they receive more when it comes from their peers.

    However, I don’t think Twitter alone is sufficient to make all that happen – more for your points about the 140 character limits and the lack of depth and memory that you cite in the comments and in your previous article. But it is a mistake to think of the tool in terms of 1 to 1 communications that can only be made efficient through automation. Twitter IS well-suited to connect many people together on almost any topic, and this is where the possibilities lie. Twitter by itself may not be enough for real support, but making Twitter more like an IVR is not going to make it better, it will only make it as limited as IVRs currently are.


  9. Great discussion, Esteban. I agree with most of what’s been said and wanted to shed light from the Customer Experience Analytics side of things.

    As far as Twitter or any other social media channel is concerned, it’s just that – another way to interact with your customers. And if there’s a way to log it and track it properly you can then integrate it into the complete cross-channel view of your customer experience. Was Twitter the last resort? Did this customer try the website first, then the IVR and only then decided to openly voice his frustrations for everyone (including your competition) to see? Are you aware that the way you respond to this tweet might be viewable to everyone?

    There obviously a lot to take into account when analyzing this new form of customer interaction. I don’t believe it’s mature enough just yet, but with new developments by several companies (CoTweet had some good news yesterday) it will eventually grow into something that really can’t be ignored.

    IVR is a slam dunk. Route people to the right resources, let them self-serve if they want to, let them speak to a live person at any point in the experience, etc. Social media channels are a little tougher to gauge. How do you get notified? Is it reliable enough? And of course, there’s the whole issue of information security.

    I’m very excited to see how all this develops. Thanks for the great topic.


  10. Great discussion!

    I agree to an extent. I’m a fan of using Twitter for customer service; I’ve written about it a few times, myself:

    But you’re right–Twitter won’t replace your existing IVR or any other type of service. Instead, we need to look to expanding our service options horizontally. Some folks want to call on the phone, so let’s improve the IVR. Some want an instant Twitter response they can rave/complain about after, so let’s monitor social media channels. Some want to meet us at trade shows.

    You get the idea. Customer service is about using all available channels to serve and reach out to our customers.


  11. Esteban,

    A great conversation with many different facets and perspectives. If you don’t mind, I would quite like to bring another angle in and compare Twitter to an outbound voice dialer – this was prompted by the megaphone comment which which amused me and actually prompted me to comment on your blog.

    My simple point is that rather than considering Twitter to be like an IVR; which as you correctly point out are generally poorly implemented and not liked by customers. I think you could also consider it to be like an outbound voice dialer providing notifications to your customers.

    Now I realise that these things are generally not liked by customers since they are essentially an intrusion or there is nobody there to field the call when they dialer calls you. However, given the broadcast SMS nature of Twitter and the fact it is non-invasive in the way that a dialer just isn’t – Twitter is never going to interrupt the TV for example then there are merits to be had in thinking about it in this way.

    The other thing to consider here, is that if you are a company with a Twitter presence then you are opting in to what that company has to say. Therefore, one could argue that you are reaching high value customers simply by virtue of the fact that they must be loyal to your company in brand to follow you in the first place.

    Personally, I quite like the 140 character limit, it makes people be more concise which in a world where there is a lot to be said and time is precious is quite welcome.



  12. Esteban-

    Twitter is a good place for IVR systems, both for customer service and marketing. I think Wim was accurate when mentioning that, at scale the customer experience will require automation. I don’t agree with Wim that once the automation is in place Twitter as a channel won’t have advantages over other channels. The light-weight experience is my first choice for “most” customer service and as tools are developed to enhance that experience this sentiment will grow. Yes, we build those tools.

    Brands won’t have the option of just listening. As Esteban keeps pointing out marketing should be the primary focus for most brands on Twitter. It won’t be well received to market without listening over a channel that is two-way.



    1. Dave,

      I am actually very curious at this point to see where twitter will go in the future with versions and functions before committing to it full time. We all know it has flaws, but if they are fixed they could be riding the wave for a long time. If not, or not fixed on time, I will be curious to see what happens.



Comments are closed.

%d bloggers like this: