However, most marketers can, and do, measure the monthly rate of growth using their own numbers as a benchmark. If you’re currently doing this, I’ll show you two new ways to measure the growth of your list using data from third-party competitive intelligence tools — so you can be not only best in class, but also be better than your competitors.

1. Benchmarking List Growth Against Your Competition

You’ve grown your list 26% since last year. You pat yourself on the back, and your management team commends you on a job well done. Unfortunately, your competitor grew their email list by 65%.

In 2012, Caribou Coffee had a nearly 7% larger email list than their competitor Einstein Bros Bagels, according to Return Path (my company) data. Looking at year-over-year growth tells a different story, however. Einstein Bros Bagels grew its list 65% in the following year, and now has a subscriber list that’s 20% more than Caribou’s.

One could chalk this up to growing pains; but, when looking at the percentage of total emails read, Einstein Bros Bagels achieves a 47% higher rate of emails read than Caribou, showing it’s not only the size of your list that counts, but what you do with it.

Caribou also sends five times more emails than Einstein Bros Bagels, likely to make up for the lack of list growth and engagement. Nearly twice as many people delete Caribou’s emails without ever opening them.

The moral of the story here is to benchmark your list growth against your competitors rather than yourself. The trends and insights you notice in your own program will make more sense when compared to your competition.

2. Shared Subscribers

Using email competitive intelligence tools [Editor's note: examples of such tools include eDataSource, Emailium and Return Path's Inbox Insight, though each has different capabilities], you can measure how much of your subscriber list is shared with your competitors. This can yield insights such as the health of your list, how many of your active subscribers were lost due to churn, or that you’ll likely lose because of churn. And, if you’re the underdog, you can determine how well you’re doing to chip away at a competitor’s audience.

I looked at two of the biggest tax preparation packages, TurboTax and H&R Block.  TurboTax may be the underdog here as far as email list size is concerned, but do they have a winning chance? Of course they do, but only if they see where their strengths and weaknesses are.

By comparing H&R Block’s list to TurboTax’s email list, I see that H&R Block has an email list over seven times larger. The more interesting part is the number of subscribers shared between the two. Five point three percent of TurboTax’s subscribers also subscribe to H&R Block. Looking closely at those shared subscribers, however, we see that TurboTax has a read rate 175% higher than H&R Block.

The “so what” here is that TurboTax is continuing to eat away from H&R Block’s subscriber base. H&R Block can look at the engagement rates from the shared subscribers to determine if they can realistically outperform TurboTax in the inbox if they want to stop this trend.

What other ways would you use competitive intelligence and subscriber lists to outperform your competition?

But, the Vitamix (a high speed blender) has shifted my understanding and approach to food. I’ve learned to use different types of raw foods to fuel whatever activity I’m about to engage in (tip: try kale for mental clarity).

This shift has improved my performance and productivity both mentally and physically (some in my family even point to a sunnier disposition) — even though I never considered my nutrition to be lacking or my performance (or disposition) to be something that was holding me back.

What does this have to do with social media marketing tools? For now, all tools pretty much fall into the Aero Press category. And, what we really need is a Vitamix.

Like the Aero Press, today’s tools help marketers accomplish things they already knew they were going to do each day, like posting to various social networks and responding to customers in social channels.

These tools also help social marketers find the answers to things they already knew they needed to know – like how to time those tweets or which customers to focus on engaging. While all that is unarguably useful, there are some things this generation of tools simply cannot do, such as:

• Finding solutions for problems you didn’t know existed
• Finding answers to questions you didn’t know to ask
• Discovering insights that will fundamentally change the way that you use/approach social media in the first place

Transformative Analytics

The next generation of social media marketing tools will do more than organize, manage and spit out pretty reports full of out-of-context facts. Transformative analytics will be at the heart of these tools, meaning that the data provided will go beyond standard metrics about activity and/or a single encounter in social channels.

In fact, this next generation of tools will be applicable to not only social media, but to all marketing. Below are three characteristics I very much expect to see in next-generation tools:

1. Broader Data Synthesis: As a (former) vendor in this space, I can personally attest to the difficulty of aggregating data from different social channels and combining it with even the most obvious sources like Web analytics data. And, this is only the tip of the iceberg in terms of the data that actually needs to be aggregated.

Truly transformative analytics will require combining data from sources beyond social channels or even online sources, including data from multiple corporate databases and a very long list of third-party sources.

2. More Information On People: We’re not going to get better at marketing — particularly social media marketing — until we get a lot better at knowing more about people. Understanding the person behind the action (or interaction) is critical to innovative use of social media. The big issue here, of course, is privacy.

3. Better Ways To Work With Data: At a granular level, I’m talking about better visualization, more flexibility with queries, and a more robust set of work-bench features. But overall, what I think we need are tools that emphasize data and pattern discovery rather than tools that emphasize data confirmation. (Think “The Machine” but without the Orwellian overtones.)

In social media marketing, as in just about every information-centric field, the challenge will always be to develop actionable insights into data of ever-increasing volume, diversity and complexity.

Today’s Tools Vs. Transformative Analytics

Do shades of all of these things exist in some of the tools on the market today? Sure. But the reality is that we are a long way from truly achieving transformative analytics. In fact, there is a long list of significant barriers to overcome. (Chief among them: a viable business model that balances the expense of such an undertaking with the market’s preference for cheap tools on an SaaS basis.)

Reality aside, what’s on your wish list for a silo-busting, next-generation set of social media marketing analytics tools? Let’s combine our perspectives, because we’ll never build it if we don’t start envisioning it.

Image via Shutterstock

Often, you want to look at your metrics over a longer time period than usual in order to identify long-term trends. This is especially true if your metrics are “jagged” and move around a lot. You have two choices:

• You could use a larger time period. If you usually watch daily metrics, switch to weekly. Or monthly. Or quarterly. This will smooth out your data and will also smooth out directional trends emerging from your data right at the time when you want to catch such changes.
• You could use an average of values of your metrics. For example, instead of weekly period, you could examine the average over the last seven days. Since the past data examined changes every day, this is often referred to as a “moving average,” since the days looked at move forward in time as you collect more and more data.

Now, you could use a simple moving average (SMA), where the weights of each day of data is given equal weight. You might even do something slightly more sophisticated by using varying weights to the Nth day, referred to as a weighted moving average (WMA). But, either solution leaves you with two issues:

1. A loss of memory: if you’re looking back N periods to calculate your average, you’ve effectively forgotten what happened more than X+1 days ago. If something important occurred any earlier than that, too bad.
2. A data-scale problem: you have to keep the data around back far enough to re-calculate the average each day forward. Not a big deal if you’re looking at shorter periods (daily, weekly, even monthly), but this can be cumbersome if you’re examining longer trends.

Enter the exponential moving average (EMA). With the EMA, the “memory” of past metrics values is never forgotten, though it is gradually given less and less weight over time. Further, since the EMA is based on what the value of your metric is today combined with what the EMA was yesterday, you don’t have to keep large amounts of past data on hand to update each day’s new values.

I’ll show you a quick example in a moment. But if you do anything with Google AdWords, you are almost certainly concerned with something that is based on an exponential moving average: the Quality Score (“QS”). As any AdWords advertiser knows, Google reports QS scores only at a keyword level, though it’s commonly believed that Google has QS numbers for AdGroups, Ads and even entire accounts.

That’s a lot of number crunching, even for Google. And you may have noticed that when Google reports your QS to you, it’s presented as a single number, not as a time series. Thus, your QS is not itself a metric, but rather a property of your account. It’s also well known that the history of your account plays a great part in determining your QS.

Sounds familiar, right? Because those are exactly the characteristics of an exponential moving average discussed above. I don’t think I’m going far out on a limb to assert that Google is using an EMA (or equivalent) to store, score and report your Quality Score.

Why not use this same technique when examining your own metrics? Here’s how the EMA is calculated.

Let’s say you want your past data to have a strong memory and thus have yesterday’s value of the EMA be pretty important when calculating today’s value. In an EMA, one uses a factor (usually called alpha) which represents a constant smoothing of the previous values of the metric. The value of alpha is anywhere between [0,1]; the higher the number, the more quickly past data is discounted.

In our example, we want to discount past data slowly, so we’ll sent alpha at, say 1/3 (33%). That’s the weight of the current data. So, normalizing, that means the weight of the past data is (1 – 1/3) = 2/3. If the value of the EMA yesterday was, say 27, and the value of today’s metric is, say, 30, then the value of the EMA today is:

EMA (today) = (2/3)*27 + (1/3)*30 = 18 + 10 = 28

If you want an EMA that responds faster to changes in trend, then set alpha higher, so that past values of the EMA are discounted faster. For example, set alpha to 80%. Now:

EMA (today) = (20%)*27 + (80%)*30 = 5.4 + 24 = 29.4

This may give you a much easier way to track important changes to your metrics while at the same time smoothing out any spurious noise in the data as it occurs. Your mileage may vary, but if Google is doing it with your Quality Score, it certainly is a trick I’d want in my bag o’ metric wizardry!

While the relationship between bids is straightforward to understand, quality score is much more opaque, being affected by many factors such as click-through rate (CTR), keyword and ad copy relevance, quality of landing page, etc.

Many marketers have studied quality score extensively, and there are some excellent overviews on the subject. However, the specific drivers of quality score change, and how exactly they act, remains largely unknown.

Below are a few facts that will hopefully help shed some light on the inner workings of this mysterious metric.

1. Quality Scores Can Rise Or Fall Rapidly From Day To Day

Below is aggregated Google advertiser data showing how often quality score changed by a certain amount on a daily basis. Although larger shifts tend to occur less frequently, there are still a good number of cases where the quality score moved by 5 points or more from one day to the next.

The distribution appears linear on a log scale, where the difference in frequency for each unit change in quality score is around 40%. This means that if there are 100 keywords with a 1 unit shift in quality score during a period of time, around 40 keywords will have shifted by 2 units, 16 by 3 units, and so on.

Quality score change is not always stepwise or incremental: depending on the account and competitive landscape, there are cases where quality scores shift from one extreme to the other.

2. Some Quality Score Shifts Bounce Back Immediately

A good portion of quality score shifts last only temporarily. Of all keywords that saw a quality score change, 21% bounced back to the former score within one day, 32% within two days, and 39% within three. This means that even if you implement a change and see a quality score shift for better or for worse, it is best to wait a few days before arriving at a final conclusion.

3. A Significant Shift In CTR Can Trigger A Rapid Change In Quality Score

After analyzing data on what metrics were likely to trigger changes to quality score, the two that I found to be significant were CTR and average position.

As expected, when CTR increased for a certain keyword, quality score was also likely to rise, in some cases reacting as soon as the following day. However, seeing a higher average position for a keyword without a rise in CTR increased the odds that the quality score will fall.

This shows that when Google factors in CTR to calculate quality score, it models the expected CTR by each position. If recent CTR for the keyword or creative is above Google’s projected curve, the quality score will adjust upwards. Similarly, falling below Google’s projections has the potential to negatively impact quality score.

4. Quality Score Is More Volatile On Head Terms Than Tail Terms

For most metrics, greater volume generally means more stability. Not so for quality score: this is a metric in which head terms change more frequently and by a larger amount than on tail terms (excepting strong brand keywords that are locked in at a high quality score).

A possible explanation for this trend is this: if Google’s quality score algorithm waits until a certain threshold of significance is reached before making changes, then more traffic will allow the threshold to be reached faster.

When trying to gauge the effect of a landing page or ad copy test, it may help to focus on the head terms, not only because of the larger impact, but also because they will be a stronger and faster directional indicator of quality score change than mid or tail terms.

5.  A Unit Change In Quality Score Affects CPC By 5%.

It is well known that a lower quality score means that you have to pay more per click to maintain a given position. The distribution of quality score is relative to competition, so the actual effect of quality score on CPC will be different for every advertiser.

For the accounts I looked at, a quality score difference of 1 had a 5% effect on average CPC for a keyword when controlling for bid and average position. Though this may seem insignificant at first glance, the effect is cumulative, which means that a 5 point difference in quality score can result in a 30% difference in CPC.

Quality Score can be impacted by many different factors, many of them being qualitative and only allowing assessment via extensive trial and error.

While this process can become frustrating, some variables such as CTR, CPC, and average position are immediately measurable. By analyzing how quality score behaves with respect to each of these metrics, we can better assess impact and make faster and more effective decisions on how to improve it.

Online display advertising, however, is a different story. People go to websites for information, entertainment and engagement with other people, not to click on ads that send them elsewhere. Only 16 percent of people click on display ads in a given month, indicating that most visitors to a typical advertiser’s site get there some way other than clicking on display ads.

Optimizing for display success is still a worthwhile endeavor. In fact, less than 4 percent of consumers’ online time, less than three minutes per hour, is spent on search. The other 96 percent is spent doing things like reading email, browsing content, watching videos and engaging on social networks.

That 4 percent is invaluable to advertisers, of course. Search accounts for about half of the spending on online advertising in the United States and the United Kingdom; and in some other markets, this number is even higher.

Why Clicks Can Miss The Target For Display

Given the nature of online content consumption, how can advertisers better understand the value of their display investment and optimize accordingly? In search, optimizing clicks is essentially optimizing for conversions because they’re so closely aligned. With display ads, if you optimize for clicks, you get… more clicks. But — and this is important — display ad clickers are often not your converters. We know this because the display clicker audience profile is remarkably consistent, regardless of the campaign.

To illustrate the point, the chart below contrasts the profiles of clickers and converters in two disparate product categories — an online retailer and an insurance company. In both categories, clickers skew toward older and younger Web users. The clickers are much more similar to one another than they are to converters, the people who ultimately make purchases in their respective product categories.

Marketing Is Not A Straight Line

While click-to-conversion marketing has its defenders, display is most appropriately seen as a touch point that stimulates consumer interest along the path to eventual conversion and purchase. The figure below compares a company’s conversions (sales) attained directly through clicks with those that came sometime after consumers saw the ad (view conversions).

There were an astounding 100 times more view conversions for the campaign, with more than 23,000 view conversions and nearly 300 click conversions. Rather than clicking and immediately buying, ad viewers waited, only going to the website later to make a purchase.

This pattern is common for paid search ads, as well. Analysts have noted that, on average, it takes 3.5 days for a customer to convert after an initial click on a paid search link.

Marketing is definitely not a straight line. What does this mean? If you optimize a campaign for click conversions, you’re optimizing to influence the 300 people who clicked through. If you optimize a campaign for all conversions, you’re optimizing for the 23,000 people — the viewers and the clickers combined — who eventually get to the website and make a purchase.

How To Optimize For Conversions

If, like most advertisers, your objective is conversions rather than clicks, you can optimize your display campaign to capture them. Here are four steps to get you started:

1. Understand the characteristics of your click-based and view-based audiences. Clickers may not be buyers.
2. Measure view-based conversion volume as well as click-based conversion volume. Buyers are likely to take another look, and perhaps another and another, before deciding to purchase.
3. Judge your campaign by the conversion rate rather than the click-through rate. Or better yet, use effective cost-per-action (eCPA) to measure the effectiveness of your entire advertising investment, no matter how you’ve chosen to buy it.
4. Develop and apply a multi-touch attribution approach that includes multiple channels. This involves building a model to better account for the impact of all of the ads that touched your customers along their path to purchase and can help you overcome the shortcomings of click-based and last-touch attribution (attributing all credit to the last ad seen).

Optimizing for clicks is a proven and effective approach for search advertising. But, relying on display clicks to drive conversions is often a dead end. As we’ve illustrated, the consumers most likely to click on display ads are often vastly different from your best customers.

Therefore, if you’re optimizing your campaign for clicks, there’s a good chance you’re actually “anti-optimizing” for sales. Instead, optimize toward your ultimate objective — the campaign conversion rate — rather than toward the click-through rate and focus on developing the necessary systems and skills to understand the true impact of your ad investments throughout your customers’ path to purchase.

But, as I’ve pointed out earlier, social listening data is just the tip of the iceberg. The variety and volume of data available to marketers keep getting bigger — and more interesting — each day.

Agencies that know how to harness and use social data creatively will profit and grow; those that do not will be left behind.  Below are a few prominent reasons why social data proficiency is a critical competency for the agency of the future:

Multi-Channel Success

Social data will hold the key to insights that lead to optimization — not just of social media initiatives, but of marketing in other channels and across multiple channels, as well. (As a simple example, consider the use of trending topics to inform content or campaigns — but the possibilities here are endless.) In fact, at the Forrester Research eBusiness Forum, 80% of respondents said they plan to use social data more in 2013 to inform campaigns in other channels.

Understanding & Predicting Buyers

Increasingly, customers will expect you to know them. Social data can play a pivotal role in building a deep understanding of customers (and also of prospects at various stages of the buy cycle).

This sort of knowledge will be absolutely fundamental for all marketing of the future (i.e., giving people what they want, when they want it and understanding consumers well enough to develop persistent relationships and predict behavior). Without it, the fall-back option is old-school marketing tactics whose efficacy is already waning quickly.

Client Demand

According to the InfoGroup study cited above, 70% of top global marketers plan to spend more on data-related marketing initiatives in 2013; and, as noted, many companies hope to expand their exploitation of social data. But, while many tools now available are good at collecting data and producing machine-based analytics from that data, it still takes a human to tell you what to do with the data. Clients will turn to agencies and other vendors in order to apply social data to a variety of marketing activities.

Competitive Advantage

When I hear the words “social data,” I am reminded of a friend who recently took up golf. His assessment? “It’s really easy to do poorly and really hard to do well.” So, too, with collecting social data and aggregating it into a form that allows it to be sliced and diced — and integrated with other possibly heterogeneous but relevant marketing data.

It is a time-consuming and resource-intensive process. (As someone who has built a social analytics platform from scratch, I can speak to this first hand.) More importantly, analyzing social data and finding the nuggets of wisdom that drive action requires deep knowledge and a wide range of fairly rare skills. Agencies that can master these complexities and offer their own “secret sauce” will be exceptionally hard to compete against. Moreover, this type of offering will be difficult to replicate by competitors who arrive late to the game.

Credibility

We live in a data-driven world. Having the analytics chops to provide clients with answers will build trust. Facility with social data analytics will come to be a measure of an agency’s maturity and sophistication. Unique services offerings built around proprietary tools and algorithms will allow adept agencies to leverage and grow existing customer relationships, while also competing effectively for new accounts.

As consumers increase their personal and business activities in social channels — and leave a trail of data behind them about those activities — the data and sources available will continue to grow and become more complex. Becoming proficient in social data is not simple — it takes time. But, agencies that master the challenges and opportunities will come out on top.

What is your agency or company doing to become more social data proficient?

A negative metric is not necessarily something you want to get less of. For example, one might mistake “reduce bounce rate” as a negative metric — but really, it’s just a more convenient way to refer to the positive metric of “increasing the un-bounce rate.” Rather, a negative metric is something you look at to ensure that, when you have success with your positive metrics, you aren’t penalizing yourself elsewhere. I covered this earlier when I discussed the value of nothing.

If conversions go up (a positive metric) but qualified leads go down, you want to know this. If the qualified leads go down with lesser impact on the company than the conversion rate going up is worth, then the company might decide that’s okay. However, if the quality of the leads drops such that you’re losing more revenue than the increased conversion rate is worth — well, then the company might decide that the increased conversions weren’t worth it.

Or perhaps your company is trying to decrease bounce rate (a positive metric, as described above) with an eye toward boosting ad views on your pages. You might achieve that, but inadvertently cause other problems that detract from your ultimate goal. You can’t know these things if you’re not measuring them, right?

Real-Life Lessons In Identifying Negative Metrics

This post focuses on thinking about negative metrics in real-life situations: what to look for, what direction they should move in if you’re doing well or doing poorly, etc.

A great way to practice thinking about negative metrics is when you yourself have a negative customer experience. After all, you might as well as turn it into something positive, even if you’re not the one in charge of measuring. I’ll use the anecdotal story that inspired this article. This one is easy since it’s about the cable company, and cable companies are always great places to learn what not to do.

As we go along, think about your own customers and their experiences with your website, your marketing outreach, your call center and your retail presence (if any). Can you identify negative metrics you should be tracking?

I’m betting most readers will recognize the pain point in my anecdote: the other day, the cable company managed to call me — get this — seven times. Seven. Each call was about getting me into a different package of services. Negative metric #0 (so obvious it isn’t even worth counting as #1): “How many times per day/week/month do we contact the customer without his contacting us first?”

Negative Metric #1

The first time, I was happy, because I watch very little broadcast TV and wouldn’t mind hearing about a way to save money. Who wouldn’t? After confirming that I am who I am — which begs the question, “Why are you calling me if you don’t know who I am?” — the agent asked if she could put me on hold “for a few seconds” so she could look up my account and find what offers I’d be eligible for.

Now, my first reaction was, “Why are you calling me and then looking up offers for me? Why not have the offers ready to go as soon as you reach me?” But a few seconds seemed trivial enough, so I said yes. After 60 of those “few” seconds, I hung up.  Negative metric #1: “Of the total time spent with a customer on the phone, what percentage of that represents hold time?” That’d be an interesting metric, wouldn’t it?

Here’s a more subtle one: “What percentage of hold time comes from getting a response to a customer question, versus what percentage comes from us (the company) asking to put the customer on hold?”

You don’t want to penalize your call center staff for answering customer questions, but you do want to penalize the call center (systemically; not the agents) if the system itself causes the agents to put a customer on hold for information that should be readily available. So you’re looking at these negative metrics in light of the positive metric: “customers who took advantage of an offer we contacted them about.”

Negative Metric #2

An hour later, another agent called me and started the same spiel. This time I interrupted and asked to be put on the no-call list. No problem, apparently — except that the astute reader will note, “Wait, you said they called you seven times!” You needn’t be a genius to realize that being added to the no-call list didn’t take. So, negative metric #2: “How long does it take you to stop contacting customers after they tell you to leave them alone?” If your answer is “up to 30 days to take effect” then, congratulations, your company is a corporate stalker.

If I can ask them to shut off my cable and they can do that in an hour or less, surely they can flag my account for no calls in approximately the same period of time. What’s the negative metric here? “After a customer requests us not to call them, how many times do we do so anyway?” And, “Of that group who received calls after asking to but put on the no-call list, what’s our churn rate for those who got one extra unwanted call? Two? Three?”

Measuring this results in a learning: “What’s the correlation between stalking our customers and their complaining/stopping/reducing their avg $ value to us per month?” Everyone reading this knows the individual pieces making up these metrics exist — good companies put it together so they understand their negative metrics and work to reduce them; bad companies never even put the puzzle together.

Negative Metric #3

Third call, same thing. Except this time the agent was someone who spoke two different languages, but was fluent in neither. Negative metric #3: “What ratio of ‘hang-ups’ or ‘customer-didn’t-do-what-we-wanted’ occurs for native speaker agents calling our customers versus our ‘multi-lingual’ agents?” If you don’t have a problem, then this metric should be approximately equal across both groups.

Fourth call. Now I’m getting into the rhythm and really starting to think up ways to torture this process. I wait through the “several seconds” (apparently in 2013, “several” and “78″ are equivalent) it takes for the fellow to put me on hold and return with offers, and he tries to get me into a Triple Play (this is a package in which they provide cable + Internet + digital phone service). As a bundle, they can offer this at a lesser price. I don’t need a home phone (mobile only, here!) so I’m not really interested in this package, even though it’s maybe $10 cheaper per month than what I’m currently paying.

The agent seems surprised I’m not interested in saving $10 and rattles off to me the “value” of the services which, if purchased separately, would be far in excess of what I’m currently paying. Nice try, but all you’re doing is convincing me that your prices are already inflated, since you can discount them so dramatically. Kind of like the Persian rug business down the street that has been going out of business for 12 or more months. Jack up the price and then discount it heavily so people think they’re getting a bargain. It also inspires a test: “if we offer customers less of a discount, do fewer of them convert?” You’d think so, but if the discount is so large that even Grandma realizes she was over-paying to begin with, this test might turn out to be a surprise.

Negative Metric #4

“What about packages just for Internet + cable?” I ask. Oh no, no specials available for me since those are the services I already get. Negative metric #4: “What churn rate do we have for customers who find out that they have been paying far more for staying with the company than they would as a new customer?” This sort of bad impact on brand is poisonous and hard to measure.

I don’t know how you’d keep me from finding out, but surely you don’t need to throw the negative consequences of my loyalty in my face. Perhaps you could set up a test on a (very) small percentage of the customers where you explicitly say this to them and then look at the churn rate of that group compared to the control — where you don’t mention on a sales call with an existing customer that new sign-ups always get a better deal.

Wouldn’t it be an important thing to know? What if it turned out that churn among existing customers were 2-3-5x higher when they were told it was cheaper for them to discontinue service and then call back in a month and get a sweetheart deal? Wouldn’t understanding this impact how you’d go about future pricing, or at least modifying the sales script of call center agents?

Calls 5+6+7 were close repeats of the above, and I did get more negative metrics identified — which would be great if I worked at the cable company and gave a crap. I will say the final call convinced me to drop cable TV altogether, so now I just have Internet. I’m pretty sure that wasn’t the positive success metric they were going for when they first started calling me. Their negative metrics completely obscured their positive metrics. But, are they aware of this? Are you aware when this is happening at your company?

Given the commercialization surrounding the winter holiday season, this is generally a very tempting time for advertisers. With so much consumer interest in procuring the perfect holiday gift, it is little wonder that the season drums up a lot of advertising interest.

But, what of other holidays sprinkled throughout the year? Many of them also have commercial aspects that could be a marketing draw: costumes for Halloween, chocolate and flowers for Valentine’s Day, and travel — for many holidays — to bring people together to celebrate. But, are marketers currently capitalizing on these opportunities?

Optimal Times For Ad Campaigns

Examining key performance indicators around the holidays can provide marketers and advertisers with insight about the optimal times to focus your ad campaigns. With the Easter holiday so recently past, I took the opportunity to examine whether the type of holiday in question has an impact on KPIs.

Easter may not be the biggest or most hyped of holidays, but the plethora of pastel and candy sales surrounding it indicate that it is still a time when many people celebrate — and make purchases.

To quantify this, Chitika Insights took a look at the days leading up to and immediately following the March 31 holiday, from March 14 to April 1. Specifically, this study analysed Thursday through Monday of each week, looking at samples of tens of millions of online ad impressions across the Chitika Advertising Network.

Easter Holiday CTR Trend

The data used in this study is plotted as an index where the highest value of a day’s CTR is set at “1,” and all other values are plotted as percentages thereof. See the graph below for details.

Interestingly, the Easter period showed no significant change in user CTR compared to prior weeks. CTR does change by day of the week (and tends to be higher on weekends, typically); however, the shape of the trend for Easter week does not appear to be very different than the shape of the trend observed in prior weeks — implying that Easter did not greatly impact user CTR.

In comparison to other holidays, Easter has relatively little marketing surrounding it that would merit strong advertiser focus. There is commercialization of Easter, to be sure, but its main thrust of chocolates and candy is fairly low-end — the type of goods that can easily be obtained at any grocery or drugstore.

Holiday Trends Help Focus Your Advertising Efforts

In contrast, holidays like Christmas and Valentine’s Day have relatively high-end goods associated with them: more expensive presents, luxury chocolates, jewelry, etc. Easter, on the other hand, is not marketed as a holiday where luxury goods are a traditional or necessary part of the celebration; and as such, people have fewer online shopping needs associated with the holiday.

Keeping holiday trends in mind can help marketers decide where to focus their efforts. Given the relatively low commercialization around holidays like Easter, there may not be much incentive for marketers to target them.

Even so, worthwhile opportunities may still exist in certain verticals, such as travel deals to bring family and friends together for the holidays. Marketers would do well to focus their campaigns and advertising efforts across the entire spectrum of holidays, keeping in mind which ones are most relevant to their industry.

While there are some commonalities between companies in similar industry verticals, my experience suggests strongly that there simply is no “cookbook” for social media measurement that applies universally.

Image via Shutterstock

What this means is that you can’t simply watch what the gurus are measuring (or the competition, or that guy you met at the conference) and adopt it as your own. 

To make matters even more complicated, there is an enormous amount of data available about social media – and only some of it will be valuable to you. Sorting through all of it to find the bits that will be useful simply isn’t an option.

These challenges often lead to two common disasters: Analysis Paralysis (nothing gets measured and sound bites like “how can I measure the ROI of my mother” are bandied about in meetings) and/or Mindless Reporting (a standard set of useless reports are distributed regularly but no one bothers looking at them).

Both will have serious, long term consequences in the long run, especially in environments where:

1.) Tactical decisions need to be made quickly by people who manage programs

2.) Senior executives demand proof of business value for every dollar spent.

Does anyone in social media operate in an environment where these conditions do not exist?

Below are three steps that will help you avoid this fate and create a measurement approach that is effective, actionable and customized to your unique needs:

Step 1: Document Function, Audience & Business Goals

Social media serves many different functions in an organization. For example, Twitter might be used to conduct product research, to support customers, or to execute a viral marketing campaign designed for lead generation.

In each of these instances, Twitter is serving a decidedly different purpose. It is likely being used by different stakeholders or operational groups (PR, marketing , etc.). Even more importantly, it is being used to target completely different audiences.

Clearly, what should be measured will be different in each instance. In the viral marketing campaign, for example, you may want to collect data on retweets of the campaign by influencers in a particular niche. To find out if the campaign moved the needle among prospects, you’ll need additional data beyond Twitter activity.

In the customer support example, Twitter metrics might include things like inbound customer support tweets and response rate. Again, you will need additional data (beyond Twitter activity) to judge whether the customer support initiative is leading to increased repeat sales or other business benefits.

Create a spreadsheet that documents the various functions being served by social media. Note the audience being targeted (this will help you make metrics more specific). Next, link each entry in your spreadsheet back to the appropriate business goal (this will help to ensure that you include metrics that can be linked back to the big picture concerns of senior executives).

Step 2:  Identify Action Drivers

What you will do with the data you collect? Make a list: which metrics might prompt you to take an action or a series of actions that impact something that is a very high priority?

Here is a simple example: Let’s say you are spending a significant amount of budget on paid Facebook initiatives. Understanding which posts are being shared the most organically can help you quickly decide which ones are worth paying to promote. The ability to quickly identify (or be alerted to) posts that have higher than average sharing could prompt you to take immediate action that could improve your paid results significantly.

At the very least, thinking through the most important actions that will impact things that are very high priority can help you make tough resource decisions by identifying what is absolutely necessary and what would be nice to have, but perhaps not crucial.

Step 3:  Connect Data Sources

The data that is necessary to create useful, actionable metrics, as well as metrics that speak to senior executives, come from many different sources. If you really want to measure what matters, you will have to do some work to connect those sources. (Yes, this involves IT; I’m very sorry.) The only question is how much work will be necessary.

Once you have an understanding of the metrics, you need (Steps 1 and 2), to conduct an audit of the tools and systems you are currently using that hold the necessary data. This will include things like CRM systems, web analytics tools, listening platforms, marketing automation tools, email management systems, and SMMS tools (social media management systems).

Next, look for the low hanging fruit. What sources can be connected with minimal effort? What connections would help you create high-value metrics? Once you understand what needs to be connected and the level of effort required, you can make a long term plan.

The sad reality is that no tool on the market today (including ours) will give you 100% of the data and metrics you need. About the best you can expect is a set of tools that help you advance some distance down the road. The trick is to fully understand your unique needs before you start down that road.

What do the letters mean for metrics optimization?

• I is for Insight — learnings you derive from your continuous optimization efforts
• E is for Earnings — which is presumably why you’re going through the optimization process
• C is for Conversion — which is how you’re measuring the efficacy of your past I on your current E

Metrics Optimization

Think about it this way: every time you perform a test — heck, it doesn’t even have to be a formal test — you’re using your past experience, learnings and insight to guide your decisions as to what you should be testing.

And, this is not a science lab, where one might do a what-if to nudge out deep metaphysical rules of the universe. The reason you’re doing the test at all is because there’s something about what you’re proposing to test that your experience tells you might be better than what you currently have. No one on the business side does an experiment to try to make things worse (though from some of the testing schedule proposals I’ve seen at companies, I’m starting to question that!).

And, the actual improvement your company nets from your optimization efforts, your Earnings, come after you’ve applied your Insight. Fair enough, especially on the net… because everyone takes credit when there’s a net increase in revenue.

You know the old adage, “success has many fathers, but failure is an orphan.” So, when you propose a test, perform it well, and it results in a net positive to the company — anyone who’s touched any part of effort is surely going to be in line for some of the credit.

Note: If you’re at a big enough company, you may well notice that the total amount of credit claimed is often far in excess than the total amount of effort put in. Ah, but I digress…

Rewarding Smart Failure

What happens when your testing efforts don’t result in a net gain for the company — the test fails? Well, that’s precisely where you actually drive future Insight.

Your learnings always come from your failings. And even better, after a failure, there isn’t a whole line of cousins thrice-removed claiming they were instrumental to the test, competing with you for the new learnings. You often have this insight all to yourself.

This is a very human sort of occurrence, going back to the first caveman who put his hand in a fire. At several of the large companies I consult with, I actually encourage them to publicly reward folks who propose tests that don’t pan out — assuming the test ideas weren’t idiotic in the first place — because the insight that comes from great test ideas that fail is the entire basis for the great test ideas that succeed.

In other words, ”I Before E, Except After C” = “Insight comes before Earnings, Except after Conversion improvements.”

Great companies embrace failure because it makes them smarter. How does your company reward smart failure?