You know hashtags can extend your reach, right? You know you can get more shares, more traffic and more followers, right? So you toss a few hashtags into your tweets and Instagram posts based on the most obvious words… STOP! Hashtags can be very powerful on Twitter and Instagram. And LinkedIn is even pushing hashtags, so who knows? The power of hashtags on Twitter is visually evident on this graph. But doing research pays huge dividends. We recently prepared tweets for two different divisions of a huge organizational client. We did the research. And we worked with each division to write the tweets. It was beautiful. A tale of two tweets The first division ran the tweet as we had presented it. This was a small division, with a social media presence of 30,000 followers on two Twitter accounts. We were very happy with the results: 1 comment 77 shares 138 likes The second division handed the tweet over to its in-house social media team. Argh. They butchered the tweet by changing both the text and the hashtags. Furthermore, they created their own image and suppressed the Twitter card we had programmed into the page. They sent it to over 400,000 followers on three Twitter Accounts and one FaceBook account. In theory, the second division should have received 13 times the engagement and 13 times the traffic to its web pages. Well, here’s what it got: 4 comments 70 shares 175 likes Why was the first tweet so successful? In a word, research. Here’s what we did: We started with the words we thought most people would use. We plugged them into Google Trends to compare them. We brainstormed for more terms and ran comparisons. When we were sure we had the most-used search terms, we created hashtags from them. Remember that hashtags are single words. So “accounting jobs” becomes “#accountingjobs”, for instance (let’s use this example, because the real words would reveal the client and violate confidentiality). We went to Hashtags.org to see what similar terms people were using. It’s easy to do. You just type in a hashtag to see the tweets where it is used. Then you look at what other hashtags are used in those tweets. The more “niche” the community of users is, the more useful these results are, because the publishers of the tweets are also the consumers of the tweets. With a semi-refined list, we headed over to Hashtagify.me to get a tag cloud for each potential hashtag. Once we had our final list, we had some culling to do. It is generally not a great idea to load a tweet with more than two or three hashtags, or else it looks “cluttered and messy”, as Erik Emanueli puts it. However, we had a couple three-letter acronym hashtags, and we were feeling frisky. We culled to five hashtags, because a couple of them worked at the end of the text anyway. How to evaluate a hashtag When researching hashtags, remember that numbers don’t tell the whole story. Just because a lot of people “use” a hashtag by placing it in their posts, doesn’t mean that a lot of people “use” the hashtag by following it. You don’t care how many fellow tweeters are broadcasting the hashtag. You care how many twitter users will see your post because of it. Hashtag rules of thumb Here are a few rules of thumb we used to decide which hashtags would be more useful in reaching new readers: Avoid very general hashtags, such as #books or #jobs or #business, which are so general that nobody would follow, even if millions of people “use” them in their posts. The audience for these terms is just too wide to be used for discovery or search, and even if somebody did use them, our tweet would be lost in a huge pile of others. Ask if there is a community of interest around the hashtag. There are some types of people that soak up whatever information there is on a topic from fellow users. Hobbyists are like that, such as amateur historians or classic film buffs. For anniversaries, for example, we found that a lot of people follow #OTD and #OnThisDay. It was critical for us to identify for each hashtag a fan base. A hashtag is only useful if there is an identifiable community of interest. Check how many people retweet posts that include the hashtag. We ignored CNN and CBC tweets when we did this. Of course, those get retweeted just because the sources have huge audiences who tend to retweet whatever they post. The hashtags probably didn’t play much of a role in driving retweets. Instead, we looked at smaller accounts. If posts by Joe and Jane average with a hashtag get retweeted several times, that is a good sign of a strong hashtag. It means they probably got exposure beyond their own audience as a result of the hashtag. What else can you do? We had already set up the Twitter card, but that’s another tutorial for another day. And we had written text that was short and sweet, that also contextualized what was on the Twitter card. In this case, the context was why we were tweeting it on this day (an anniversary). It could have been why we are sharing in on Twitter. Or it could have been to tie the content to current events. We also added a couple emojis, which called attention to the tweet and also helped separate the hashtags from other text. We didn’t just throw emojis in for fun. The two emojis we added gave important visual context to draw the eyes of our target market. They showed visually the country and the sector . You can’t do this kind of research for every tweet or Instagram post, unless you work for a huge organization. But when you have something important, it’s worth doing the research. Even looking at a single tool to see what is being used (published) the most, then thinking about what is most likely to actually be used (followed and/or searched) will put you miles ahead of just winging it.