Why campaign analytics are a crystal ball, not a rear-view mirror
Few marketers are big fans of post-campaign campaign analytics. After the campaign is complete, the creativity that they injected to kick it off in the first place has been spent. The excitement that they started the campaign with is over.
And the rush of success — or the horrible taste of failure — is past.
Reviewing marketing campaign analytics, however, can be incredibly valuable.
You learn what went wrong. What went right. You see how your target audience reacted, or how you activated a totally different audience that you didn’t even think the campaign would impact. You learn something new about your creative, and you also learn something new from almost every campaign about the channels and media that you’re using to execute the campaign.
As good as it can be to measure campaign effectiveness, your campaign analysis should not just be about the past.
More than anything else, it’s about the future.
What’s past is prologue
William Shakespeare knew a lot about marketing. (Seriously!)
After all, you don’t get to be the most successful playwright in English history by not understanding audiences and messaging. He probably would have been a great ad writer too: Shakespeare contributed more pithy phrases to the English language than anyone else.
But even the bard of Stratford-upon-Avon himself probably didn’t know he was saying something profound about marketing when he wrote four words in Act 2 of The Tempest:
“What’s past is prologue.”
Campaign analytics are necessarily about the past (let’s not get into predictive analytics just yet!). And that’s OK. But the point of campaign measurement and ROI analytics is to guide marketers to the future.
So they need to be used as prologue … prologue to newer and greater heights of execution, growth, and success.
Review: what you’re typically doing in campaign analytics
Traditionally, campaign analytics have mostly used in post marketing campaign analysis.
You review your strategy and objectives to refresh what your goals for the campaign were. You run through your creative strategy, and then you look at campaign reporting by channel, breaking down your results to identify, in detail, how the campaign performed.
That includes analyzing metrics that are generally channel-specific, so a multi-channel campaign that includes organic email outreach and paid mobile advertising will have multiple metrics for each of the included channels. For example, email will have open and click-through rates as two of your key metrics. Similarly, mobile marketing would have cost per install or CPM as a key metric.
Finally, you’re going to summarize your findings and provide recommendations for future updates.
Modern marketing: High-volume marketing and “campaigns”
But modern marketing works a little different. And it requires a different analytical strategy.
Modern high-scale marketing will definitely utilize any or even all of those steps, but marketers might do them in different order, or in different ways.
Modern marketers aren’t running one campaign at a time. Often, they’re not just running several, either. Many of these marketers are managing spend on the order of tens of millions of dollars a month, and they’ve got multiple continuous campaigns with no specific beginning or endpoint.
Some campaigns are intermittent; pausing occasionally.
Others morph into each other. Few are discrete and separate.
Even those that are specific — a Christmas campaign, for instance — might actually be more marked by a change in creative than traditional elements of an entirely new campaign: different targeting, adjusted channels, new partners, completely new adsets, new creative, new offers, and so on.
With the switch from discrete campaigns to multiple ongoing efforts, campaign reports and campaign analysis is changing as well.
Campaign reporting: from episodic to continuous
If you have fewer discrete campaigns but more continuous activity, it shouldn’t be shocking that campaign analytics follows the same trajectory.
Modern marketers don’t sit down and “do analytics” for a number of hours or a day when a campaign reaches and endpoint or conclusion. Instead, they “do analytics” for a few hours every week (or even every day) on major media sources and campaigns in progress. They make tweaks and see what the impact is.
(Also, while there’s definitely still manual work to be done and some Excel hell, or Tableau terror, marketers expect their analytical packages to do much of the work for them … and are impatient when they don’t do it all).
For marketing scientists, though, there are few campaign-finished-let’s-analyze-performance moments in time.
Instead there’s just continuous additions of creative to feed their media partners’ algorithms’ appetite for continuous testing and tweaking, plus incrementality testing to ensure they’re achieving lift that can be isolated for all of their efforts and each of their partners.
Campaign analytics in Singular
So how do campaign analytics work in Singular? Well, there’s far too much to tell in just one post, but here’s a quick overview of some of the capability. (Also, see more here.)
In an overview look at your marketing, you’re essentially looking for a high-level summary: what’s happening, what’s going right, what’s going wrong. Here’s an example dashboard which gives a fictitious client an overview of their marketing activity for a particular week:
Note that the time period is adjustable, as are the KPIs you want to see.
We’re not looking at a particular campaign analytics here: we’re looking at multiple campaigns and all ongoing activity. The goal is not to get too deep just yet. You’re getting the big picture, and you can click into obvious outliers (like the bottom blue line in the chart) later to get more details.
Campaign analysis (channels, partners, results, other dimensions)
If you’re looking to dive deeper into a campaign, that data is at your fingertips in the reports section of your dashboard.
Select what you’re looking for, and then you can also choose metrics, custom dimensions, and even custom events to modify your view of the data:
Additionally, you can track campaign changes to see what impacts your results the most. Tracking campaign changes is incredibly powerful because in an era when marketers are constantly tweaking campaigns to optimize performance, you need to know both when something works and why it did for you to be able to replicate it in the future.
(This is fairly unique; expect more details on this from Singular soon!)
Perhaps you’d like a more proactive approach to metrics. In this case, the campaign measurement tool allows you to set goals and then track your progress to them.
Marketing results on all your ongoing campaigns (and, if you have them, episodic ones) is great. But sometimes you want to know what creative — or types of creative — works best across all your marketing.
Marketing campaign analysis is not complete without getting insight into creative performance. Creative analytics allows you to find that out … and then optimize for your best-performing creative in any individual campaign. In this case, we’re looking at creative clusters (clusters of similar creative) by CTR and conversion rate, among other things.
It’s pretty obvious that customer retention or mobile user retention is critical for success. Pouring or 99 users out of the bottom of your acquisition bucket every time you acquire 100 makes acquisition campaigns very expensive.
Even saving a few percent of your users could halve your cost of acquisition. So understanding why users churn, which users churn, and what makes users churn is an extremely high-leverage use of time … and campaign analytics.
And that makes a bit of time spent studying retention reports and figuring out strategies to improve them time very well spent.
There are some metrics in each vertical that are obvious and common to all marketers in that sector. Sales, for instance, in retail, plus cart size. Installs, for example, plus in-app purchases in mobile gaming.
Most of those will appear in your campaign analytics screens by default.
But each sophisticated growth team develops its own understanding of the mechanics of their apps. And their own custom metrics that highlight critical areas of the customer journey. If you’re in ride-sharing, you might have a metric around cost per first ride. Or, you might learn over time that a calculated metric of cost per first 10 rides is more indicative of long-term success, and therefore a better gauge by which to measure user and customer acquisition campaigns.
For these, Singular allows you to set up custom metrics. The result is that you save huge amounts of time when checking your campaign analytics because it’s already in your language and your
Cost, revenue, ROI, and ROAS
You can’t calculate ROI on marketing or ROAS on advertising unless you also have complete and solid cost data and revenue data. Both are critical components of Singular’s campaign analytics.
Cost is a fairly deceptive thing. It seems easy — how hard can it be to total up how much you’re spending on different partners, channels, and platforms?
But then you realize that discrepancies are rampant, reconciliations are almost impossible, and you are only getting cost in multiple reports from different sources in different formats, none of which is aligned 100% with the actual results (conversions) you’re driving from those sources. It’s not standardized, and it’s not normalized.
How can you use that to calculate ROI and ROAS?
That’s why the most sophisticated marketing brands on the planet trust Singular to aggregate their cost: partners like Lyft, Rovio, AirBnB, Ubisoft, and dozens we can’t mention.
The right way to do it is with complete integration into each media source on both the payment and fulfillment sides. Payment is where you set up campaigns, bids, and tactics. Fulfillment is what you get: installs, conversions, sign-ups, purchases. Getting both sets of data and combining them are critical to generating truly useful (and accurate) ROI.
Automation might not be something you think about when you think campaign analytics. But don’t you want cost and conversion data pulled and combined automatically? Don’t you want alerts on out-of-band events like extremely low conversion rates, or high spend levels, or other things like that. Automation — and ever-increasing amounts of it — can only help marketers.
There’s a lot more we could say. And we’ll add to this over time. But as the famous saying doesn’t really go, a demo is worth a thousand words.
Isn’t it time you give Singular a chance to show you what we can do for you?