Back in October, I watched the U.S. presidential debates live on television with about 67 million other Americans. TV was the main event, despite the fact that platforms like YouTube and Ustream offered some great non-TV alternatives. But like any other big media event these days, there was a significant online conversation happening in real-time. And for the most part, that was a conversation between and among humans. But every once in a while, a robot got a word or two in.
Twitter, which was my go-to platform for my two-screen political experience, generated more than 10 million debate-related tweets. But the tweet that caught my eye -- no, I didn't read all 10 million #Debate tweets -- wasn't political. Not even close.
Adam Kleinberg, the CEO of Traction, made a keen observation somewhere in the middle of the first debate. While the majority of us on Twitter were dissecting the political theater in real-time, a few people had clearly left their robots in charge. And those unattended robots were doing some real damage.
"What I found ironic [was] that some of the biggest 'gurus' of digital and social marketing were the ones revealing themselves as robotic posers in the social sphere," Kleinberg wrote in a blog post the next day that called out some of those gurus.
But while the debates exposed some social media gurus as clumsy, at best, and charlatans, at worst, I began to think about the larger role of bots in social. After all, a lot brands use bots, especially on Twitter. But I wondered if it was really a good idea to use a bot at all?
Yes, you can use a bot
First, the good news for bots: In some cases, bots can be quite helpful.
"Bots can be an invaluable scheduling tool that lets an agency or a brand automate tasks that don't need to be run through human hands," says Justin Oh, senior digital strategist with 22squared. "But bots are just a tool, and like a lot of tools, they can easily be misused, which happens a lot in digital."
According to Andrew Cunningham, a community manager at HUGE, bots can serve a few important marketing functions. But he cautions, "Overall, bots are a means to an end. They should be used infrequently, if ever, and are more of a business tool than an effective means of social reach."
- Bots are cheaper than people. That may not matter across the board, but some brands are understaffed.
- The technology that allows you to post pictures and videos on Facebook and Twitter can be "finicky," so some marketers prefer to upload and test in advance and use a bot to schedule the post.
- Bots make it easier to target a message by time zone.
When robots are left unattended
Frankly, we've all seen a robot screw-up. Those screw-ups are as common in social as hashtags and typos. But from a brand perspective, bot-related screw-ups can be a disaster, showing "extreme insensitivity or a total lack of awareness," says Cunningham.
For Cunningham, the NRA's Twitter account in the wake of the Aurora, Colorado movie theater shooting is a good example of bots gone bad. But he says that it is just one of many examples of a bot that was preloaded with a message that turned out to be insensitive by the time it actually posted.
Now, there's a perfectly good reason for that tweet. The message was totally innocuous and on brand when it was written and loaded into whatever bot the NRA uses. But when it posted after the shooting in Colorado, it looked like the NRA was either clueless or mean. In reality, it was probably more like careless. Because, while Cunningham says there are some good reasons for deploying bots, he's also adamant about keeping humans in the loop.
"It always makes sense to still have a set of eyes checking each post once it does go live to ensure there are no issues [such as] cancellations, change of plans, national event, etc.," says Cunningham.
But beyond adequate deployment and supervision of bots, their use raises some serious questions about the nature of social.
What your bot says about your brand
"In a lot of ways, how a brand uses a bot is a litmus test for how committed they really are to social," says Oh.
But it's not just as simple as saying, "Brands that use bots just don't get social," because there are obviously places where brands are going to use bots effectively. It's really about whether or not the brand understands that social has changed us as consumers, and, by extension, changed the way we think about brands.
"There's been a huge reset in terms of how consumers see brands," says Kleinberg. "Consumers now have an expectation of humanity from brands."
That doesn't necessarily mean a brand needs to put specific names and faces to its social profiles -- there is, after all, a big risk that those people will leave the company eventually and take that personalized brand equity with them. But it does mean that we want a little more human touch from our brands, especially when it comes to social.
What works for one brand won't necessarily work for another brand, of course. But each brand should experiment with appointing social media spokespeople or just try striking a different, more personal tone online.
For Kleinberg, it's a little like the automated telephone menus that have become so common when you call a large company. Sure, those menus can walk you through a lot of options and provide good information, but they shouldn't be a model for how your brand communicates online, especially on platforms like Facebook and Twitter where the expectation is that everyone is -- or ought to be -- a human.
"Just be human," says Kleinberg. "I think too many brands obsess over content curation and that gets them into trouble -- whether they're using bots to post or not -- because they're not thinking about creating something that a human would actually want to see or read."
And if those brands are using a bot for a significant portion of their messaging, they're really not talking at all, says Oh. What those brands are doing is a lot more like walking into a crowded room, turning to face the corner, and shouting nonsense at regular intervals.
"Nobody wants to be carpet-bombed with clunky messages," says Oh. "That's not what I'd call being part of the conversation."
The brands that do it well
While there are a number of brands that have gotten praise for their Twitter accounts over the years, the fact of the matter is that we're currently seeing some brand categories join the conversation, while others are lagging behind.
"In general, I'd say that airlines and banks are getting pretty good about actually talking on Twitter," says Oh. "They may use bots sparingly to push out a particular message or offer, but their accounts are manned by humans who do an excellent job of reaching out to customers when the conversation warrants it."
Admittedly, a lot of that outreach is about customer service, not marketing. And it's no surprise that brands that are especially prone to customer service nightmares have gotten particularly hands on -- in a human way -- with social media. But for all brands, no matter the category, we're way past the point when they should be debating whether or not to be in the social space. The fact of the matter is, all brands are already in the social space because their customers are using social. So the only choice brands have is how they're going to engage thoughtfully via social media.
Some brand marketers talk about engagement in social, while others talk about listening. But for Kleinberg, it's about presence, which is something you can't really get from a robot.
"If we're talking about a big brand, you really do want to have a person on that Twitter account whenever you're delivering a product to your customers," Kleinberg says.
Obviously, for airline brands that can be a 24/7 commitment. And clearly, that's not something all brands are capable of. But it all comes back to that human thing, for Kleinberg.
"The brand can set the expectation for engagement," Kleinberg says. "People understand the concept of office hours, and you can tell them when a human will be in to respond to their Tweets."
On the other hand, Kleinberg warns, if brands deploy robots on a fulltime basis, the expectation is that there's always somebody home for the brand. And that can be a really bad thing because no brand wants to be in a position of failing to meet consumer expectations.
Understand your platform
For the most part, we've talked about bots in the context of Twitter. But the truth is, bots can also be used on Facebook and several other social platforms. And while nobody I spoke with is suggesting that brands make widespread use of bots, it should be understood that the rules about using bots change slightly with the platform.
The fact that bots are a common part of the conversation on Twitter is a little surprising to Oh.
"Twitter is about one-to-one conversations," says Oh. "And it's impossible for your customer to really talk with a bot."
By contrast, Facebook does offer a little more room for brands to use bots, especially when it comes to scheduling a post.
"Bots can be really important if you just think of them as a scheduling tool for your brand's Facebook page," says Oh. "The fact is, each brand is different when it comes to scheduling the optimal time for a message, and bots allow us to both schedule and get good data on that scheduling."
While I've been semi-active on Twitter for several years, I can honestly say that I didn't get serious about it until 2012. I got serious about Twitter for a rather selfish reason -- promoting my novel.
I mention this, not to plug my book, but to illustrate just how important it is to really use the tools that are out there. It's one thing to poke around on Twitter for a few hours, read some articles, and maybe even talk to some experts; but it's something else entirely to use Twitter and understand it in the same way that you understand your community or neighborhood.
That's why Adam Kleinberg's tweet was so striking to me. He gave voice -- ok, text -- to a sentiment that I had felt for a while.
Even before I saw Kleinberg's tweet, my publisher had asked me why I wasn't using HootSuite to make more efficient use of the time I had allocated to promotion. The answer, quite simply, was that I felt funny about pre-loading my tweets. Sure, I could write a lot of tweets that would be relevant (possibly even effective), but it wouldn't be the same as talking with the people that make up my Twitter community. In fact, it wouldn't be talking with them at all. It would just be blasting them with my own messages that have nothing to do with the conversation. And while I have a pretty liberal follow policy, I always unfollow people and brands that outsource the conversation to their bots.
Michael Estrin is a freelance writer.
On Twitter? Follow iMedia Connection at @iMediaTweet.