How Twitter Bots Taught Me about Competitive Intelligence

May 16, 2017 Kyle Harper

A robot looks at the camera

The internet is an amazing place. It’s arguably the greatest marvel mankind has produced, connecting the whole globe, providing innumerable resources, and improving quality of life in countless ways.

It’s also home to a new breed of offense, grift, and general unpleasantness. So it was no surprise to me when I logged into my company’s social listening platform ready for some customer and competitive intelligence research and discovered a slew of at-mentions from a number of unsavory bot accounts promising pretty much all bodily vices in one-hundred-and-forty characters or less. While higher education brands are certainly interested in research, this field didn’t quite hit the mark for our audience.

I set about reporting the accounts and posts but noticed something interesting as I worked. Each of the posts from this bot seemed to be using a rudimentary formula that pulled language from our YouTube channel content. That was curious—why would they do that? I dug a little deeper and read through the posts from earliest to most recent. Sure enough, each post was populating titles from our YouTube channel in order of popularity starting from highest to lowest.

On the surface, it might appear like a weak tactic (our videos about faculty research, popular course offerings, and how to make a strong resume didn’t exactly resonate with the bot’s smut), but in practice it revealed some surprising truths about the way people and marketers sometimes mindlessly approach content. What if a member of my team had retweeted the post carelessly, only seeing the title of a familiar video? What if a user stumbling across the bot account confused their posts as shares from us?

What if pulling YouTube titles is actually an effective way to measure and hit a target audience’s interests?

Robot in a forest

Image attribution: Martin Reisch

The Bot Dilemma

Let’s get something out of the way to start: when it comes to marketing tasks, bots don’t really work. As marketers, we like to tell ourselves that this is because bots don’t stop to think about how to craft meaningful, personalized experiences for their audiences. In practice, I think it’s more likely due to a piece of wisdom I heard from our IT director once: “The day a grammarian gets angry and writes a phishing email, we’re all screwed.”

The bots that people regularly interact with on forums, in YouTube comment threads, or on social media lack a particular level of crafting or clout in many cases. This hasn’t stopped brands like Coca-Cola from looking into how they might use bots to build ads of their own, albeit with a higher level of sophistication than most of what’s online today. But given Coke’s previous trouble with bots as a marketing tool, it’ll be interesting to see if this is a gambit that actually survives once the general consumer populace catches wind of it.

Rather than looking at bots as a piece of marketing technology to use, I decided to think about marketing bots from the perspective of the coder behind them. If I had to make a program that would do Facebook, YouTube, Instagram, or Twitter research, pull a single piece of data, and then use that to drive interest, what data would I look at?

Cardboard robot

Image attribution: Janko Ferlic

Non-Robotic Competitive Intelligence

Your brand, no matter what you do, has at least one competent competitor (if you don’t think you do, you just haven’t found them yet, and you should be worried). As a competitor in your space, competing for the same or similar audience, they have to post, share, and circulate all sorts of content the same way your team does. This material can be a powerful source of information for your team.

While no replacement for more expensive and thorough market research, there are a lot of interesting insights your content team can gather from competitor material. Here are just a few ways you can “think like a bot” to improve your content strategy:

  • Video Platform Review: Just like the bot I saw at work, reviewing a competitor’s YouTube or Vimeo feed is a great way to get ideas for aesthetic, themes, information, and production styles your audience cares about. But beyond just what’s popular, YouTube in particular also makes it easy to do sentiment analysis on how people are reacting to popular material. Knowing how your audience might react to a particular combination of creative choices is a powerful way to drive specific emotions for specific occasions.
  • Keyword Density Analysis: Many SEO/SEM platforms include tools that help you analyze what keywords a competitor might be trying to hit (for those who don’t use this particular marketing technology, you can also use free tools like this one). This is somewhat trickier information to work with, as your team will have to make tactical decision about whether to attack, defend, or abstain from those words based on competitiveness. It’s still good information to have, and doing similar keyword density research against user comments can sometimes yield interesting results. Plug a long list of comments into an analyzer, and you might be surprised what phrases come up repeatedly that you wouldn’t have thought to research for your keyword strategy.
  • Talk to Other “Bots”: While bots aren’t technically popular in the marketing world today, automation is. While the difference between the two is largely semantic, some of the same principles still hold. (If your competitor is willing to automate a particular message or piece of content, this signals that they believe that message or content is likely to resonate most if not all of the time with their audience.) So if you haven’t chatted with your competitor’s online customer service, signed up for email drip campaigns, or tested for auto-responses from any number of site interactions, then you’re missing out on a powerful space to learn what sales propositions, experiences, or language really resonates with your contested audience.

The web today is crowded with too many bots for marketers to start acting robotically as well. The work that content marketers do to research, craft, and curate excellent experiences for their audiences will continue to outperform halfhearted automated messages. But by giving yourself the space and time to think like a bot about your competitor’s content, you’ll likely find there’s low-hanging competitive intelligence fruit ripe for your team’s taking.

Subscribe to the Content Standard

The post How Twitter Bots Taught Me about Competitive Intelligence appeared first on The Content Standard by Skyword.

About the Author


More Content by Kyle Harper
Previous Article
Healthcare Marketing Is Pioneering Value-Based Content, and the Rest of the World Should Pay Attention
Healthcare Marketing Is Pioneering Value-Based Content, and the Rest of the World Should Pay Attention

Content not connecting with your readers? When creating value-based content, take a lesson from the healthc...

Next Article
Mary Mazzio Takes a Stand: A Conversation about Advocacy, Authenticity, and Brand Storytelling
Mary Mazzio Takes a Stand: A Conversation about Advocacy, Authenticity, and Brand Storytelling

Documentary filmmaker Mary Mazzio discusses the push for authenticity in brand storytelling and the crucial...