Does Influencer Marketing Actually Work? A HubSpot Blog Experiment

I ran my first experiment in influencer marketing in middle school. My mom had bought me a pair of practical (her words) but hideous (my words) sneakers -- and insisted I wear them to gym class.

Believing -- maybe rightfully -- that my reputation hinged on having cool footwear, I convinced my mom to buy a second pair as a gift for my friend Kelly's birthday. Kelly was the coolest girl in sixth grade. Everything she did turned into a trend.

Kelly wore her sneakers to gym, probably at behest of her mom. Next thing I knew, everyone was rocking my ugly shoes. I was trendy, not lame. The experiment was a total success.

I didn't think about influencer marketing again until I joined HubSpot. On the blog team, we're always looking for ways to scale traffic -- which gets continually harder as your audience grows. We've already captured much of the "low-hanging fruit." Partnering with influencers could help us reach new readers while bringing our current ones fresh insights and promoting worthy thought leaders. Win-win-win.

When I started experimenting with influencer marketing strategies for the HubSpot Sales Blog seven months ago, I was operating under a few key assumptions:

1. If one influencer is good, 23 is better.
2. A huge name on a standard quality blog post is better than an unknown name on a great quality post.
3. Partnering with influencers is an efficient and scalable way to grow traffic.

Turns out, those assumptions were mostly -- even completely -- false.

Experiment #1: Is Influencer Marketing a Silver Bullet?

Let's go back to November 2016, when I set out to create a blog post that received 10,000 views or more in its first 30 days.

For context, average views per Sales Blog post for November was 3,180.

I decided to create an influencer round-up (a type of post gathers quotes from multiple influencers on a single theme or topic). Harnessing the reach of some of the biggest names in sales through their social shares would surely push the post over the 10K line.

Since there are relatively few sales influencers, and their expertise is pretty varied, I chose a broad theme for the round-up post: How to Have Your Best Sales Year Yet. Each month received its own section and covered a different aspect of selling, from qualifying to giving demos.

Then it was time to get the influencers on board. First, I identified 30 sales experts and thought leaders, and wrote a personalized email to each one.

Twenty-three influencers agreed to participate. That's approximately 77% of the 30 I approached -- a great response rate that was likely bolstered by the personalized emails.

Next, I wrote custom questions for each influencer based on their area of expertise. I collected their interview answers for the post through a combination of phone and email interviews. I also asked each influencer to commit to promoting the post on LinkedIn and Facebook, as well as their email newsletter if they had one.

After the post went live, I created custom tracking URLs for each influencer to use for their share(s). This step allowed me to see exactly how much traffic every individual was responsible for. I also created a custom image for social media each influencer with a quote from their interview:

From start to finish, the entire post took me roughly 18 hours. A standard, non-influencer post takes me 45 minutes to an hour. That's a big difference in time.

The Results:

The post ended up getting around 9,100 views in the first 30 days. The influencers were responsible for 4,143 of those views -- each sharing their link on Facebook and Twitter within three days of the launch date, with several sharing the link on LinkedIn as well.

Sounds like our influencer strategy worked, right? Well, not quite.

When I dug deeper, I discovered the influencer traffic wasn't distributed evenly. In fact, one person was responsible for 77% of all influencer traffic.

Some influencers drove less than 20 views each.

These results killed my first assumption: If one influencer is good, 23 is better.

Experiment #2: Is Targeted Influencer Marketing a Silver Bullet?

With this in mind, I had two goals for the next post in my influencer marketing experiment:

• Feature as many "heavy hitters" as possible.
• Feature the influencers' products -- giving them a clear incentive to promote the piece beyond being quoted as a thought leader.

I chose The 20 Most Highly-Rated Sales Books of All Time as the theme, since it would enable me to satisfy both criteria. I sent an email to each author with a custom URL and a polite request for them to share it.

The Results:

All in all, the influencers drove around 5,700 views. Guess who was responsible for 60% of those views? The same influencer who was responsible for 77% of the traffic for the first post.

It's not surprising this influencer is such a reliable source of traffic. He has more than 1.7 million Facebook fans. A mere 0.002% of them clicked on the link to the HubSpot Sales Blog.

What is surprising is that the other influencers are driving such little traffic. It takes roughly the same amount of time to get a quote from the other influencers as this superstar one, but the results are completely disproportional.

Experiment #3: Looking for a Scalable Solution

Is the solution to concentrate solely on massive influencers? Unfortunately, there aren't enough out there in the sales space.

What if we skipped the time-consuming process of picking a theme, finding experts, reaching out to them, developing questions, transcribing and editing their answers, and putting everything together into a post? What if I started with content that already existed, cutting out the creation process entirely?

I started handpicking influencers based on their network size and asking to publish their original content and/or repurpose some of their existing content -- along with a social share, of course.

This strategy requires much less time. Each post would receive fewer page views, but I'd be able to produce more in the same amount of time -- meaning we'd drive more overall traffic.

It's also great for the influencers. They've already done the hard work of content creation; now they can sit back and enjoy its amplification. We send their posts out to our 50,000-plus email subscribers and give them the option of adding in-line links and a CTA to their website or virtual offers.

I started this experiment with an excerpt of a book from a well-known author, who has a sizable audience.

The Results:

The post got 2,143 views -- and 1,200 of those came from the HubSpot Sales Blog email subscribers. Because the author didn't use the tracking URL I sent her, I can't say definitively how many views her Facebook and LinkedIn posts generated, although it's likely around 300. Not that impressive.

We see similar results whenever we publish posts for "the name" versus the content. If the writing isn't relevant, insightful, or helpful, it doesn't seem to matter who's got the byline -- the post tends to strike out.

There goes my second assumption: A huge name on a standard quality blog post is better than an unknown name on a great quality post.

Experiment #4: Content First, Influencer Second

With that in mind, I decided to focus on content first, influencer second. I looked for sales experts with unique perspectives and ideas and largely ignored the size of their audience.

A sales leader I found on LinkedIn is a perfect example. He's not a recognizable name, but he's built a solid following (almost 10,000 followers) by consistently posting entertaining, helpful articles on LinkedIn.

His posts we've published get an average of 4,600 views.

Tom Niesen, CEO of Acuity Training, belongs in this category as well. He wrote a post about upfront contracts that generated roughly 6,000 views before Sandler Training shared it on social and drove 1,000 more.

These partnerships might not create one massive traffic spike, but they're lightweight for me to manage, generate content I'd be happy to share with our readers no matter who wrote it, and are a scalable way to work with influencers to produce content that consistently performs above average.

My third assumption: Partnering with influencers is an efficient and scalable way to grow traffic -- still true, if it's the right influencer.

3 Influencer Marketing Takeaways

It turns out influencer marketing has gotten a bit more complex since middle school. After running these four experiments, here are my three main takeaways.

1) If you're trying to drive traffic, use your time and energy to get one major influencer on board rather than five to 10 mid-level ones.

An influencer's impact depends on both their audience size and engagement -- you can get similar results from an influencer with tons of followers and low engagement and one with a medium following but crazy-high engagement.

I recommend looking at the influencer's average Facebook post performance. If they're receiving 600-1000+ reactions and 500+ shares on any given post, they're probably a major influencer. Validate this by giving them a dedicated tracking URL to share and seeing how much traffic they drive.

2) That being said, quality still matters.

Weak content rarely performs well even with a "big name."

When evaluating content for the Sales Blog, I ask myself these questions:

• Are the ideas relevant?
• Have we covered this material in other posts?
• Does the author give enough explanation, detail, and instruction that the reader can immediately apply the concepts?
• If they make a controversial argument, do they sufficiently back it up with research?
• Is the post engaging from beginning to end?

If the answers aren't all "yes," then I'll ask the author to edit the piece.

3) Prioritizing content leads to posts that get traffic on their own merits.

I'm proud of our readers. They're not star-struck -- if a post is great, they read it and share it, regardless of the byline.

Ultimately, our best "hack" for growing traffic? Publishing great content. That's an assumption I'm happy to operate under.

How does your content team approach influencer marketing? Let us know in the comments.