hopes for good metrics

Last week, I posted the following update across a couple of networks:

Please don’t use “Reach” as a viable metric. It’s a BS number to get clients to pay more for data: don’t go down that road

The update resulted in some great discussion on both sides, with agreement and counterpoints on how viable reach is as a metric.

These included:

Lets not be so quick to disregard Reach. After all, big companies buy Superbowl spots exactly for the Reach it gives them. – Dino Dogan.

Reach has 2 parts. How many people saw something (actual reach) and how many potentially could see something (fictitious reach). If I have a reach of 2 million but at the time I post every day only 3 people see the post what is the real reach? 3. – Howie Goldfarb.

Reach is definitely ephemeral because reach doesn’t really matter. Action does. Click throughs, downloads, sticky site traffic, purchases, that kind of thing is what really matters. – Shelly Kramer.

But what about quantifiable reach? Meaning, reach is tracked against response & metrics? – Mitch Arno.

All good points – yet (for me personally, anyhoo), I’d disagree on Dino and Mitch’s points on when reach could be used as a viable metric. Here’s why.

The Attraction of Reach

By definition, reach is simply a potential number. It can be broken down a few ways:

  • On social networks, reach can be the amount of people that may interact with your content, through either seeing it directly, or through one of their connections, or even a social ad;
  • On blogs, reach can be the amount of people that interacted with your post, through direct readership, a shared link, or a forwarded email;
  • On media, reach can be the amount of people that could see a TV show, a print ad, billboards, etc, through direct contact or shared discussion later.

Marketers like to use reach as a client metric because it sounds impressive. Instead of having to be restricted by real actionable data, reach can be used instead.

  • “Twitter User X has 1,000 followers, but a reach of 100,000. That’s 100,000 new customers just waiting for us.”
  • “Blogger Y has 10,000 subscribers as well as 1,000 followers. Even more new customers just waiting for us.”
  • Free Local Newspaper Z has a reach of 50,000 homes delivered to. That’s 50,000 new customers waiting for us.”

And so on, and so on. Which would all be great and dandy if reach was a viable metric, but it’s not.

The Problem With Reach

The reason reach isn’t truly viable as a metric is simple – it’s based on the hope that eyeballs are available.

On Twitter, for example, reach is defined by the potential audience of a user you’re connecting with. So, someone may have 1,000 followers, but the followers of these connections combined may equal 100,000 (at least).

When User X with the 1,000 followers tweets something, the hope is that person’s followers will see the tweet and reshare with their audience. If a follower has 100 followers themselves, that’s now another 100 eyeballs to potentially see the share. Potentially being the key word.

Because no-one is online all the time, just waiting for a tweet. And that’s just the live, organic stuff – when you add bots and fake followers into the equation, things become even less impressive, as we can see by the following example of the same Twitter power user.

Twitter Reach

In the image above (click to expand), I used TweetReach to gauge the potential reach and visibility a certain power user could offer. The user in question has over 217,000 followers.

As you can see from the image, the estimated reach is just short of 200,000, with potential exposure (followers of followers) at just over 2.6 million. Impressive, right?

But this is where reach as a metric falls down, as we can see in the next image.

Status People Fake Follower Check

Using the Fake Follower Check software from Status People, I entered the same user’s details, to see how many of his 217,000+ followers were real. The definition of “real” here is simple – active and not bots or fake accounts.

As you can see by the image above, there’s a huge disconnect between how many of this user’s 217,000 followers are classed as real, active followers.

Suddenly, instead of having access to a potential 2.6 million impressions (which is what the potential reach of the followers pre-fake check would give), you’re now down to 182,000. Still impressive, but a hell of a lot less than 2.6 million.

And that’s assuming that each of the 182,000 “real” numbers are around when this power user sends out a tweet.

This is why reach is such a crappy metric to be using – and yet many marketers use it to increase their rates when charging clients for audience size.

The Real “Reach” Metrics

To Dino and Mitch’s earlier points at the start of this post – I agree, to a small degree, that reach us used by the likes of the Super Bowl ad buy agencies, as well as quantifying the reach. But there’s still the overarching problem that’s it’s a hit-and-hope metric.

In the Super Bowl example, companies use data from Nielsen, one of the world’s leading audience measurement companies – but it’s aggregated data based on average audience behaviour, versus specific targeted individual behaviour. This means it doesn’t take into account people getting up during ads to grab more beer, food, etc.

With Mitch’s example of quantifying reach, this gets closer to viable metrics, but it’s still relying on the bigger number that can’t be quantified. If we move the conversation over to quantifiable metrics based on more in-depth filtering, then we can get more actionable insights.

  • Ignore the big number and target the demographic instead – male/female, locale, language, age;
  • Use search history to identify common keywords and themes relevant to your brand;
  • Use software like oneQube to identify when users are on Twitter, for instance, and who they’re talking to. Do the same with software relevant to the platform you’re interested in;
  • Collate this information with Google Analytics or similar to understand what content is attractive and what content was physically shared and acted upon;
  • Set goals based on real action (as Shelly mentioned in her examples) – click throughs, recommendations, downloads/purchases, referrals, etc;
  • Compare these actions to the target audience size to determine reach vs. actual ratio.

By doing this, you’re now beginning to bypass reach alone and get to the real nitty gritty – your true audience size, as well as the real numbers to filter and use as a proper business metric.

Reach itself can be a starting point when it comes to social proof and how many eyeballs could be reached – but only as a starting point. You need to then do the legwork to define your captive audience.

Otherwise you’re constantly just reaching for what could be – and few businesses, if any, can live on that alone.