Measurement & Metrics

Digital media predictions for a post-GDPR world

May 14, 2018 by Susan Getgood

Last week, a business acquaintance made a ballsy prediction on LinkedIn for which he took a fair amount of flak. Posting a comment, I noted that his essential point was not that much of a stretch. Of course, there were nuances to the argument, but a long statement with a bunch of caveats wouldn’t be nearly as interesting or thought-provoking. Predictions have to push at the edges or people won’t pay attention. So, I figured it was time to step onto the ledge and make some predictions about the future of digital advertising. I’d love to hear what you think, agree or disagree.

1. In a post-GDPR world, targeted advertising is going to become more expensive. Not immediately. There is still a lot of confusion (some real, some manufactured) about what is required. In the end, though, the increased costs of delivering compliant targeted audiences combined with a decrease in available inventory as consumers retract or refuse permission, will increase ad rates to premium, not programmatic, rates. I’ve been arguing this for a while and the pundits are finally admitting it.

2. Context will be queen again. In order to reach the right audience, we will turn to content-centric, community-centric marketing tactics. In part because they ARE more effective in converting customers, but mostly because the cost differential between them and targeted/programmatic advertising won’t be so great. It will be just that much easier to take the leap of faith to try these tactics, which are still working out their success metrics (more about that in a bit.)

This is the opportunity for niche publishers and brands to create content for their micro-audiences that both provides the necessary context for ads and sponsored content and creates a relationship and value exchange in which the consumer is more likely to give permission for the specific use of personal data. As an example, Pepper & Wits, a new content site for women 45+ who are navigating menopause is owned by P&G; we certainly should expect sponsored content and other brand outreach but the primary purpose is to offer value to the consumer and targeted context, not targeted ads.

3. “If you aren’t paying, you’re the product” is a cliché whose days are numbered. Consumers will start to care about privacy. It has been a LONG time coming, but I truly believe that consumers are finally understanding the true cost of free digital services/platforms, and are going to want real value in exchange for the right to use their personal data. “Legitimate business interest” is not going to be sufficient to use a person’s data without permission. Publishers and brands who are offering great content and building relationships with consumers will have a far better chance of obtaining (and retaining) permission.

4. Things are about to get more competitive. Facebook and Google are vulnerable in this new advertising economy. They aren’t in danger by any means, but their models are based on scale, not relationships. They have no friends, just users. This opens up a sliver of opportunity for niche content publishers to create better experiences for consumers and pick off a little wedge of the pie. The duopoly will still get (more than) its fair share, but they will be handicapped in delivering their biggest strength, targeted audiences at scale, so the little guys can dart in and nab their share in areas where Facebook and Google cannot play effectively, content and community. Niche publishers, bloggers, even brands who can make the long haul investment. And Amazon. Amazon already does a better job of community than either Facebook or Google, is nipping at their heels and has a distinct e-commerce advantage. I also predict Reddit will make a strong play for the “community ad dollar.”

5. All of this will drive innovation. Certainly, in ad tech to manage consent, but that’s just the tip of the iceberg. Folks are already working on challenges like better measurement and attribution models that assign value properly and proportionally to all the players in the value chain, not just the last click. This is crucial for branded content, video, social media and influencer marketing.

Community is also getting some attention. Two start-ups doing interesting things to connect brands with consumers: Social Data Collective and Suzy. Suzy (AskSuzy.com) is the evolution of social media and influencer business Crowdtap. It helps advertisers make and manage meaningful connections with customers by offering them access to its super panel of consumers. Social Data Collective has a slightly different approach; it asks consumers to share personal data with the brands they love in exchange for products/services.

But I think even more interesting things are on the horizon. For example, aggregators of consumer information that validate consumer audiences and can compare them across properties, including blogs and YouTube and all the other places that community will form, not just sites big enough to register on comScore. Audience data all the way down the long tail to validate that the context IS delivering the right audience. This “data hole” has been the bête noire of the influencer marketing business, but there wasn’t a strong enough incentive to solve it, when advertisers could just buy targeted audiences cheaply. To sell content and community as the right context, you need the metrics to prove that you’ve got the goods. It’s coming. I am certain. And it won’t need PII (personally identifiable information) to do it.

We also need tools to give consumers control over their privacy across platforms and processors. Right now, if you think someone has compromised your private data or is using it without permission, good luck tracking it down. Blockchain is the best bet for creating this privacy audit trail. Someone is certainly working on this already. In fact, if you know who, I’d love to chat with them!!!

So there you have it. My predictions for the digital media industry. If you need me, I’ll just be out here teetering on the ledge.

P.S. If you want to check out the prediction of my business acquaintance, you can find it on LinkedIn.

Yes, my friends, influencers do have influence. Here’s why we’ve gotten confused.

March 3, 2018 by Susan Getgood

Earlier this week, my friend Toby Bloomberg tagged me into a conversation on LinkedIn about evaluating influencers for branded content programs, spurred by a Medium post that argued influencers don’t really have influence.

My comment got too long. What a surprise! Not. So I am posting it on the blog in its entirety. For context:

Influencers have influence. Ultimately, influencers are your customers. They buy (or don’t) your products and services, and they talk to each other. Some influence a few, some influence many. But there is always value in customer -centric marketing.

The issue here is that we collectively have done a few things that contribute to the perception that influencers don’t have influence. I could write a book on this, and perhaps I should, in between trying to re-establish my consulting practice once again after 7 years inside.

First, we talk about one group, influencers, which is misleading. There are really multiple subgroups, each of which has different characteristics, and should be folded into the marketing plan with different strategies and tactics. Microinfluencers are your customers who individually do not have tons of followers, but in aggregate can create a volume of earned media. Mid-tier content creators/bloggers are also your customers, but they have built a larger following on a blog or social platform, and can be tapped into as content producers for both the endorsement value and their distribution channels. Celebrity influencers are much more about their scale, and somewhat less about their role as your actual customer.

Second, we need to activate these different types of influencers in ways that match their reach, influence and status as your customer. The same strategies don’t work across the board. We should start with our business objective, tap into the influencer population that will best help achieve that objective and the define a measurement strategy matched to the program. We also need to communicate those objectives and expectations clearly to the influencers we work with. It is OKAY to ask someone you are compensating to deliver a certain result. Our clients ask it of us. We should ask it of the people we work with. Not necessarily with crowdsourced strategies tapping into hundreds of microinfluencers, but certainly for content programs with mid-tier folks and celebrities. You cannot get mad at non-delivery if you don’t set an expectation.

Third, we need to ground our content programs in relationships with our influencers. After all, they are our customers. There are no shortcuts to relationships. There is no tool, no database, no algorithm that can substitute for the relationship. Use automated tools to get started, to manage content production, to monitor and measure results, to understand the performance of your influencers. But don’t expect that you can define a target audience, search a database for influencers that match the target, or whose audience does, or both, hire them to do something without the due diligence of getting to know them, and then hope for the best. If you know the influencers, you’ll know whether and where they have influence.

Finally, the original article was interesting in that it conflated two related but different marketing strategies, influencer marketing and thought leadership, in an attempt to make the case for thought leadership over influencer marketing. You can certainly tap into a customer (influencer) population as part of a thought leadership strategy, but by and large, thought leadership is about conveying ideas and building an expert reputation for the individuals and their organization, as a conduit to sales. Expertise does matter, and we may reach out to other experts in our field, with greater or lesser degrees of influence, to support and distribute our idea, but it is not the expertise of the customers. It is that of the brand and its representatives that is the focus.

Each strategy, influencer marketing and thought leadership, has its place in the marketing mix, and we don’t have to pick between them.

–

Agree with my ideas, but not sure how to get started? I can help with everything from strategy development and content creation to influencer, digital and social marketing, performance audits and presentation decks. Even better, the first hour is free. Email sgetgood@getgood.com to book your free consultation. I’ll give you some thought starters during our conversation, and we can go from there.

ROI and Influencer Marketing

August 10, 2017 by Susan Getgood

Continuing my series about the trends in influencer marketing: The rise of ROI.

In the early days, when blogs and social were new and shiny, return on investment was “squishy.” As influencer marketing matures, so too do the expectations, and the measurement models.

Measurement models are shifting from soft “potential reach” to firmer engagement models, and a better understanding of true awareness (eyeballs, lift) across all platforms, not just the easier-to-measure website ad impressions and content views. The best predictive models look at both awareness and engagement, to provide the necessary context for brands trying to decide which type of content will best deliver to the marketing plan.

How many people see the content across all opportunities – native, social, pageviews. defines REACH or AWARENESS. How many people act on it (clicks, likes, shares) is ENGAGEMENT.

Engagement rates will continue to be important, especially as we are increasingly able to link social actions with purchases through longitudinal studies like Nielsen Catalina, and foot traffic studies that can link a social visitor with a real life visitor, but these are expensive, and more likely to be used by large CPG and retail advertisers with big budgets. Not universal. Yet.

In the meantime, while we wait for the nirvana of proving engagement drove purchase — knowing whether someone who read that blog post last year, purchased it this year, we will rely on brand lift studies from Nielsen, Millward Brown and others, and first party reporting combined with original research. OpenUp is a start-up doing interesting work in the space of linking digital engagement with content to eventual purchase.

Advertisers want to understand the return on branded content, including influencer marketing, in the same way they evaluate their other advertising activities. Cost Per View is emerging as an AWARENESS metric alongside the click-through rate and the effective CPM that advertisers use to evaluate the overall efficiency of a media plan.

Re-visioning Measurement: A model for digital content marketing

When a marketing tactic is new, we tend to be a little forgiving when it comes to measurement. It is simply not possible to be first to market, and also have a case study to evaluate before you make your decision. We operate on gut, on past experiences that are similar, out of a desire to experiment with the new tactic. We monitor and measure, but it is to establish a benchmark, not against a benchmark.

As the tactic evolves and matures, however, a body of work begins to emerge. Successes and failures, near misses and home runs, all combine to give some indication of what works. And what doesn’t. We are at that inflection point with content marketing, and particularly with influencer marketing. Benchmarks are emerging left, right and center.

Problem is — many of these benchmarks are either the very simple pageview and click-through-rate (CTR) we started with or defined by the different technology platforms people are using, thus hard to compare with each other. In some cases, they measure things because they can, not because the measure is useful or relevant, a criticism I have oft levied at Google Analytics.

In addition to CTRs, content program benchmarks tend to rely on views (page, video, slide) to show reach, and comments and earned social to demonstrate engagement. At publishers that scale content through native, native CTRs get added to the mix. This is a good start, but volume based measurements don’t allow you to compare tactics with different budgets. More budget nearly always delivers more volume.

Adding to the complexity, Facebook and the other social platforms report in the context of their platform – likes, comments, shares – and are more than a little opaque unless you are the account owner. This makes it challenging for advertisers trying to understand their earned media. We can count it, we just have a harder time understanding the person who shared it.

Plus things change. Not every day but it feels like it.

We need to simplify to make the data we collect useful to marketers. Capture key points that let us understand the success of a particular campaign and the component tactics AND compare the campaign to other campaigns, the tactics to other tactics.

Simplify. Standardize.

Every marketing tactic we use has an awareness and an engagement component. We want you to pay attention and then do something. Isolate those and look at them separately to understand the performance of each tactic against its goals. You can also aggregate each measure to understand how the overall campaign performed.

Views — Awareness
While reach will always be important as a general gauge for awareness — the potential or available audience for a message — we need to move past who MIGHT see something, and evaluate our campaigns based on who actually did. To standardize across all platforms, we use views and actions as proxies to estimate the engaged audience of digital content.

Actions — Engagement
Absolute numbers are great to understand the VOLUME of your social engagement, but if you want to compare tactics, you need to use rates. This corrects for size. Our old friend the Click-Through-Rate is still strong here, and lets us compare all our tactics against a single measure. But, it isn’t the only useful RATE we can calculate.

We can also look at an overall engagement rate for a campaign, defined as Total Engagements/Total Reach.

For content, look at the content engagement rate. Of the people who read something, how many shared it with others? Or simply commented.

Content Engagement Rate = Actions/Views

For video, the video completion rate (completed views/total views) remains an important measure, but it underestimates the success of the content. Looking at the ratio of viewers who watched at least 25% of the video (or more than 10 seconds on Facebook) gives a more accurate measure of the video success. You can also look at content engagement rates for video.

Social is a bit more squirrely when it comes to standardized measurements across platforms. We have reach and engagements, but we don’t always have access to actual viewers of a social action due to platform and cost barriers. If we own the channel, we have better data, but influencer data depends on whether the platform allows third-party access, and if so, how much it costs to get and use it.

Right now, I am intrigued by content and sentiment analysis as the path to understanding message penetration on social. Because both paths — audience and content analysis — are on the pricey side, we collectively tend to rely on engagement metrics to understand results on social. We have the data, and we can efficiently compare across platforms.

I recommend looking at two measurements here:

Engagements:Followers
The ratio of earned:paid.

SUMMARY TABLE: CONTENT MARKETING BENCHMARKS

Cost Per View: Quantifying Awareness
Finally, even though we don’t have visibility into every view of our messages on social due to the walled gardens created by the social platforms, we can get to a very conservative estimate of how many people saw our message, and calculate a Cost Per View.

Cost Per View = Budget/Views

What’s a view? What goes into that side of the equation?

Pageviews, slide views, video views
Viewable native ad impressions. Regardless of clicks.
Viewable content amplification ad impressions. Regardless of clicks.
Earned social engagements. This is a PROXY for viewers that we can apply across all social platforms. If someone shared or liked or commented, we know they saw it. This will undercount, but it is a start.

Important: Evaluate Cost Per View against your overall content marketing campaign: Native Ads plus Content Amplification Ads plus Branded Content plus Influencer Content plus Social Promotion. Some tactics are more efficient at views than others, while others are stronger down funnel. Content creation will always be more expensive than promotion, but you need the content to promote. And so on.

Where do we go from here?
No single metric is the silver bullet. A tactic with a high cost per view can generate amazing engagement or be the perfect content base for a scale promotion. Or both. We also need to look at these measurements side by side with third party research that measures brand lift or foot traffic or message penetration, and layer in our actual sales results to get the full picture. But the important first step is to begin standardizing our metrics so we can compare campaign performance month to month, year to year, and isolate the tactics that are both efficient and effective.

More of what works, less of what doesn’t.

The importance of sentiment

June 15, 2010 by Susan Getgood

Recently I’ve noticed an increase in the number of articles/posts predicting that 2010 will be the year of measurement. When marketers using social media finally cross the Rubicon of Results.

While I am hopeful that this is true – that we’ll start measuring what matters, instead of simply counting things – I’m not sure we’re really there yet.

Social media measurement should be about actions, not simply awareness, and to matter, it needs to be tied to the measurements that matter in the business: sales and net revenue. Monitoring sentiment and counting friends, fans and followers is a step along the way to true results measurement. But it’s not the end game.

That said, understanding sentiment is a good first step, and useful for analyzing the impact of social media engagement on revenue. A sentiment score accounts for volume, so you can compare it week to week, regardless of changes in volume. You can then see if there are any trends in lead volume or sales that track consistently over time with shifts in sentiment. If you have an initiative to increase positive sentiment, you can track your results. And so on.

To calculate a simple sentiment score for a given time period, usually week or month:

(Number of positive comments – number of negative comments)/total number of comments, including neutral and mixed) * 100

Net positive mentions divided by total number of comments, which accounts for volume differences week to week. Times 100 to put the answer on a 100-point scale.

The scale runs 100 to -100, with zero the neutral point. The closer to 100 on the positive side, the more positive the sentiment. On the negative side, the more negative.

You don’t need any fancy tools. A spreadsheet will do. All you need is a method for capturing social media mentions across the channels you wish to measure (blogs, Facebook, Twitter etc.), a standard way to assign sentiment to each message and the counts for each sentiment.

Marketing using social bookmarking

February 14, 2009 by Susan Getgood

Earlier this week, I had an interesting phone call with the marketing person from a game portal. He had called to avail himself of my “one hour free” offer. He primarily wanted to talk about the potential of social bookmarking — such as StumbleUpon, Digg, Delicious, Kirtsy, Mixx and Reddit — as a marketing tactic.

I suspect he was disappointed in the call. Unlike an analyst, I don’t have great pronouncements about trends. What I do is discuss your business issues and make concrete suggestions for your marketing and social media plans. There is no single answer, just avenues for exploration.

Here’s what you should look at regarding social bookmarking.

Social bookmarking as a tactic is a bit like Google Ads. Easy to do but hard to do well. To be successful with Google Ads, you need to look at your stats every day. Adjust keywords. Tweak content.

Social bookmarking is the same.You need to look at your web analytics, understand what the social bookmark sites deliver to your site and be willing to tweak your content to be appealing to the ones delivering real traffic, real prospects.

Black hat tactics like paying for submissions are a non-starter. They may deliver traffic — until you are caught — but the traffic won’t convert. They won’t buy, they won’t come back.

And that’s what you are looking for. Buyers and loyal visitors.

Unlike Google Ads, though, you can get some benefit dabbling in social bookmarking. If you aren’t willing to invest the time to understand which of the social bookmarking sites are productive for you, you can still get some incidental benefit by placing the links or a widget like Share This on your site. Or by occasionally pinging your friends to Digg a post. No real investment, no expectation, no worries.

If you want to spend the time though, here’s my suggestion.

First, look at your referrers. Which social bookmarking sites are delivering the most traffic, and the most productive traffic, to your site? It is bound to be different depending on what you write, what you offer. Every social bookmarking site has a user base and possibly even a perspective. You need to understand which ones are important to you, to your customers.

Then look at the type of content that attracts these visitors. If it is consistent over time — the same type of content attracts the visitors from a desired social bookmarking site, you have guidance for what to write to attract them in future. It’s not that different from the process used in television ratings. The networks tailor their content to the most productive audience.

It’s simple marketing math.

Making social media measure up

January 30, 2009 by Susan Getgood

As I mentioned in a previous post, I led an advanced workshop on blogger relations at a local pr agency last week. A significant portion of the discussion centered on measurement which offered a great opportunity to revisit my thoughts on the topic. This post covers some of the material I prepared for the workshop.

Measurement isn’t magic.

It’s also not the same as monitoring even though the two activities use some of the same tools and we often confuse them. Monitoring is qualitative. It looks at outputs — media coverage, blog posts, microblog streams. It’s purpose is to evaluate attitudes. It’s extremely important at the outset of any marketing campaign and it can inform part of the measurement. But it is not sufficient in itself. A clip “book” and a calculation of reach (how many people were potentially exposed) is good information to have, but it only measures potential awareness.

And last I heard, no one ever went into business or ran for office to make folks more aware. The goal is to sell some product or win the election. A result.

Monitoring is “tell me everything you know.” Measurement asks specific questions. What was the result? Did we achieve our objectives?

Measurement must be based on a desired behavior or action, not attitude. Outcome, not output.

It is important to choose a measureable outcome, not some squishy thing that can’t be assessed by an action or behavior. The best measures are action or behavior: evaluate a product, intend to buy, recommend, purchase.

Unfortunately, it isn’t always easy to link marketing campaigns directly to sales and other purchasing behaviors. So we are often left with web metrics. Useful ones include unique visitors, referrers and path, time on site, and for blog-supported programs, inbound links and comments.

These indicators are better than nothing, but the key to success is to define the measurement at the outset, not as an afterthought and build it into your program.

For example, a dedicated microsite gives you a set of web metrics 100% related to the social media program. A coupon or online discount code lets you track campaign-driven sales. Even something as simple as a badge that customers can put on their own sites can provide some basic information.

The $25,000 question is, why aren’t more people measuring at this depth? Why are we still talking about awareness, not about purchase behaviors?

It’s a combination of fear and ignorance.

Let’s start with the ignorance. We aren’t asking the right questions. If you set your objective as something squishy like “raise awareness,” your measurable result will be equally squishy and irrelevant to business success. Fine and dandy if we could magically pull unlimited money for marketing programs out of a hat. But we can’t.

This is where the fear comes in. We’re afraid that robust measurement may show that all that wonderful awareness didn’t translate into actual purchase. The more money we spent on the program, the more afraid we are. Safer to stay in the comfort zone of awareness.

Except that won’t fly. Not in this economy, and really, not ever. We must be accountable for results.

We need to shift our thinking a little bit. Big programs that don’t work can be career, or at least job, ending events. No one wants to be the guy that put forth a huge social media flop.

Think smaller, think pilot programs. Test, measure, evaluate, and then scale up.

Be more tolerant of failure. Fast, less expensive failure, but don’t dismiss a marketing tactic if a program doesn’t have the initial results you wanted. Figure out why so you don’t repeat the same mistake the next time.

And for goodness sake, ask the right questions so you can know, not guess, that you succeeded.

—

Finally, a quick plug for my contest over at Snapshot Chronicles. Prize is a $100 JCPenney gift card.

RSS Feed Experiment

November 21, 2008 by Susan Getgood

Just because people subscribe to a feed, doesn’t mean they read it. So I’ve been conducting an experiment of sorts. Not particularly scientific and absolutely no baseline measure, so consider it more anecdotal than anything else.

As regular readers know, about a month ago, I moved Marketing Roadmaps to the WordPress platform from TypePad but I did not redirect my Feedburner feed. This was a deliberate choice, as I am convinced that feeds — especially feeds that have been active for four years as the original Roadmaps one was — accumulate waste circulation. This is people who have subscribed to the blog in multiple feed readers, probably serially, as they bounce from reader to reader. I wanted to stop carrying these dupes.

I’ll use myself as an example. When I started reading blogs, I used the Bloglines feedreader. When Google introduced its feedreader, I switched over to it, and then about a year or so ago, I switched to Newsgator. When I switched however I did not unsubscribe the abandoned readers from the feeds. So there are a significant number of feeds to which I am subscribed at least twice, possibly three times.

While I generally refrain from assuming that my behavior is reflective of the rest of the population, in this case, I think my pattern is pretty typical for the small 11% of the online population that uses feedreaders (Forrester data.)

How much waste is there in my old feed? A month into the cut-over, subscribers to the new Marketing Roadmaps feed are about 6% of the total number of subscribers to the old feed. Since this is totally unscientific research, we can’t make the correlation that only 6% have re-subscribed, but I will bet that it isn’t far from the truth.

What does that tell us? I make no claim that my results are indicative of anything other than my blog and its audience. However, my data hints that feedreader subscriber numbers are very inflated, especially for long running blogs, and may not be the best basis for evaluating a blog’s readership or creating ranking systems. If used at all, feed reader subscribers should not be weighted heavily.

Further substantiation. The number of visitors and unique visitors per month to the old TypePad site in August and September, the last two full months of its life as the active blog, and to the new site in its first full month are about the same. Traffic to the TypePad site is also falling off. While this is all extremely unscientific and has absolutely no statistical validity, it does support my belief that the readers who read my blog on the blog is a fairly stable number, and most have followed over to read at the new site.

How useful is my data? What can other people extrapolate? A lot or a little I suppose, depending on how honest you want to be about how many readers are actually, regularly reading your blogs. It’s up to you.

For myself, I have a great deal of confidence that I truly know how many regular readers Marketing Roadmaps has. Thanks for sticking with me. You know who you are 🙂