Blog post

Is Social Marketing Actually a Grand Illusion?

By Martin Kihn | October 24, 2014 | 6 Comments

social media analyticsMarketing

It’s time to get real about measuring social marketing. I’m not talking about ads on Facebook or, ahem, Snapchat — that’s just advertising. I’m talking about that thing we’re supposed to be doing around the clock 24/7: creating great content that reflects our values and sharing it on our social channels so whoever’s there will pass it along and we will . . . well, what happens next is never entirely clear, but it’s got to be something. Right?

Let’s come right out and ask it: Does social marketing have any real impact on any business metric?

That’s business metric. Not social metric. Using fan counts or retweets as a measure of impact is kind of like asking the Cake Boss how your diet is going. No, I’m talking about connecting all the stuff we’re doing to influence social chatter to the things that ultimately matter: revenue, market share, profitability, stock price, market cap.

Good news: Rather than offering mere opinion, here’s actual data. For six months ending last May, 2014, Gartner used a social listening tool called Tracx to collect and analyze public conversations about a select list of companies, on popular social channels such as Facebook and Twitter and blogs. The targets we chose were leading digital marketing or ad tech related software companies or digital marketing agencies. There were 28 of them. This is what we found:

There was no observable correlation between social marketing volume and business success for digital marketing software and services companies.

You heard me. Before we start parsing out explanations — or you tell me the methodology is flawed — let me take you through what we saw. Also note that this was not a peer reviewed experiment and is not official published Gartner research. It is a conversation starter.

Sample: The 28 companies in our sample included both large brands, such as IBM and Adobe, and smaller ones such as Datalogix, DataXu and Turn. These are companies that come up the most in inquiries with our clients. We also included a few prominent agencies such as Razorfish and Xaxis.

Filters: In setting up our monitors, we tried to filter out irrelevant conversation (to the extent possible) using negative matches, and to include only conversations relevant to digital marketing. For example, we were only interested in following Google Analytics, not Google’s self-driving cars, and IBM’s digital marketing platform, not its hardware or consulting businesses. Some brands such as Turn were very difficult to handle, but we did our best.

Channels: We looked at public conversations on the brand’s own pages, as well as conversations happening elsewhere. We included news sites, Twitter, blogs, Facebook, Google+, YouTube, Tumblr, Instagram, LinkedIn and others.


(1) Some brands make a lot more noise than others

Our sample could be divided into three types of brands: noisy, average and quiet. Like so (showing total posts captured):


Obviously, Google is a hot topic but it’s always an outlier. So is comScore because it comes up in conversations that cite it as a source for media metrics. On the other hand, HubSpot is an acknowledged master of social marketing — that’s what it does, after all, enabling brands to market themselves, — so it’s clearly doing something right. Among the other players, Marketo (a public company), Sitecore (CMS) and Optimizely (an optimization tool) are making some noise. On the other hand, our friends at Turn and Rocket Fuel (pre-brouhaha) and SAS are more subdued.

But does any of this matter? Well, it’s hard to see how. Everyone loves HubSpot, and it recently had a successful IPO, but is it really a better company than Turn? Better at social marketing, certainly. It’s market cap is about $1 billion. On the other hand, Marketo is only making about 20% as much social noise, but is valued at $1.6B. Could HubSpot hold back a little and see its value rise?

And what about our wallflowers: is there comparative modesty hurting their bottom lines? [x+1], a well-regarded DMP/DSP, was acquired by Rocket Fuel recently. Certainly, its social marketing was softer than its product. Would a better program have got it a higher exit price? Maybe. But what about Turn or SAS? Turn raised $80M simoleons in January and SAS just built a new building in North Carolina. They’re making barely a peep in the social space. Are their investors and customers just ignorant people?

No, I’ll be honest and say I can’t find any obvious correlation between social marketing volume and product quality, company value, profitability, or anything really except — well — social marketing volume.

(2) Social Measurement Is Really Twitter Measurement

Part of the problem here is that social measurement itself is flawed. We don’t and can’t get a true picture of the entire “social conversation” using the available tools. Much of the social net is dark matter: IMs and emails (if you count them) are invisible; as are non-public conversations and much non-text communications like pictures. The reliance on public sources has the effect of making Twitter seem a lot more important than it is.

Here’s the channel skew of our sample:


Well over half the “conversations” happening about our companies were tweets. You’ll find this in most social reports — they are overwhelmingly reports about what’s happening on Twitter. Why? Twitter is almost 100% public and available to social monitoring tools. It encourages short bursts of outrage. Facebook is mostly non-public; remember, social monitoring tools are not in your “friends” category. They are strangers to every party. Brand pages are often public, but I suspect most real unfiltered product-related commentary does not happen on brand pages. (Blogs are probably overrepresented here because our companies — digital marketing software and services players — are the subject of a disproportionate load of blog ink.)

So I’m admitting our measurement methodology is flawed — badly flawed. But it’s not our fault. A more complete picture might show us stronger patterns, where smart social marketers dominate a productive dialogue with engaged consumers. But we don’t get to hear that conversation. Nobody hears all of it. Not even the NSA.

Although I’ve blogged previously that long posts are more likely to be successful than short posts (check me here, bros), I’m pushing it. Consider this Part 1 of 2 (or even 3). In the next part, I’ll continue revealing the results of our little experiment, including these amazing topics:

  • What is “buzz” worth? Comparing HubSpot and Marketo
  • Does self-promotion actually work?
  • Is it more important for companies to have a good product or a good social marketing program?

And so much more . . . see you next week.

Comments are closed


  • Jim Duffy says:

    Your global statement that “There was no observable correlation between social marketing volume and business success for digital marketing software and services companies” I believe misses the mark. My reasoning is that although it is difficult to measure success in this area, the social marketing field is at the precipice of dramatic change. For example in the social selling realm, Linked in used in conjunction with tools like InsideView, and smart marketers are creating paths directly to decision makers that align a companies value proposition with a organizations goals and objectives. This process also micro-targets decision making executives personal visions along with the companies they represent objectives in order to have highly focused and effective campaigns that contribute to companies top line goals. Although the social marketing efforts have not provided the results as advertised I believe like any evolution Social Marketing is in its beginning stages. Better efforts to triangulate information with “Big Data” will further drive the effectiveness of measuring the output and its correlation to the bottom line. This is from a person at the ground level looking up. I am in sales and see the difference effective social marketing campaigns can be executed in increased sales. Social Media is at the threshold of change. It has only Just Begun… I think thats a song!

    • Martin Kihn says:

      Thanks for the comment – I think part of the confusion comes when we all have slightly different definitions of “social marketing.” (Also, to clarify, my statement about correlations only applied to the specific experiment we ran, not to all social marketing.) For example, reaching out to prospects directly on LinkedIn — while it happens on a social channel, of course — seems more like traditional prospecting (like direct mail or calling or emailing) than social marketing. But that’s a definitional issue.

  • marc meyer says:

    Hi Martin, As someone who is pretty knee deep in the space, I have to take these types of posts with a grain of salt. It really does come down to mapping back social media marketing activities to ROI. The problem is when we lose the crumb trail of measuring the effectiveness of volume. Which in the social space can often happen. You lightly touched on this, but we lose the ability to measure the effect of social marketing volume when we are not privy to Twitter DM’s, convos behind Facebook’s firewall, direct conversations from Linkedin connections or blog comments that lead to direct leads from a blog post.

    Beyond the metrics that we can and cannot see, it would see somewhat premature to throw the baby out with the bathwater. The space is evolving rapidly and yes social marketers still don’t know how to separate signal from noise, but maybe we just have to get better at measuring the right things and or saying the right things at the right times?

    As the measuring and parsing of structured and unstructured social data becomes more effective and granular, we’ll “get better” at measuring the right data and quit being so enamored with likes, follows and favorites. To me, I find it really hard to discount 300,000 million users in a network or a billion in another and say “social marketing” doesn’t work. It makes for good searchable and indexable content that can be found on a blog which could turn into a business engagement, though…:)

    • Martin Kihn says:

      Thanks Marc – hopefully my second post (which tempered the hyperbolic headline here a bit) brings us back to reality. My point was just that volume does not equal impact. Or I suppose you could say: Quality over quantity. Which we should all know but too often forget.

  • Steve Casey says:

    Interesting analysis, especially the HubSpot/Marketo comparison. I hope you also include some analysis of each company’s actual financial performance as a factor in their valuation. Just checked and Marketo grew revenue 64% in 2013 to $96m and HubSpot grew 50% to $77m in the same period. So Marketo has more revenue and is growing faster than HubSpot, which are probably more significant factors in the valuation difference between the two companies than their social volume. Which relates to the previous comment — need to be careful about the definition of social marketing, if in this case it appears to be limited to the brand-awareness efforts that are easily measured on Twitter and excludes any sales-oriented digital/social marketing programs such as Twitter tailored audience campaigns that are more closely tied to demand-gen and revenue contribution but quickly convert to dark social channels.

    • Martin Kihn says:

      Thanks Steve – agreed. There’s a definition problem here but the observation doesn’t necessarily apply to B2B contexts. My second post (Weds.) tempered the conclusions a bit.