This Tactics + Insights guide, sponsored by Wiland, examines how brands and agencies use data to optimize campaigns. Marketers improve campaign performance and drive revenue growth by combining a holistic understanding of past campaigns, careful audience selection and in-flight campaign adjustments.
Agencies and brands face increasing pressure to quantify campaign performance and generate meaningful revenue. As a result, campaigns are often rapidly built and launched, rushed to completion before they can be fully optimized and then quickly forgotten in anticipation of the next campaign.
However, savvy marketers know they are more likely to hit their ambitious revenue and brand awareness KPI targets by investing in more thoughtful pre-launch analysis and post-launch observation.
“Digital marketers have busy schedules, and they’re almost always running multiple campaigns for multiple clients,” said John Goodpasture, vice president of digital agency and platform sales at Wiland. “It’s challenging to find the time to fine-tune audiences and reach the right people, but by working collaboratively with a data partner, marketers can implement changes to make campaigns more successful.”
In this Tactics + Insights Guide, Digiday and Wiland explore how successful agencies and brands work with audience and data partners to select better audiences, optimize campaigns in flight and apply past learnings to future campaigns. As they do so, silos collapse, and advertisers reach increasingly relevant consumer groups with greater effectiveness.
Assessing what worked — and what didn’t — in past activations is a vital precursor to planning a new campaign. To this end, transparency among stakeholders is essential for media teams compiling actionable insights for their future campaigns.
For agencies, this means obtaining and analyzing data from a client’s campaign history to understand better what led to successful outcomes and produced revenue. Media teams at brands or agencies — especially those responsible for multiple distinct marketing channels — must share results among stakeholders and determine how various decisions impacted one another in a given campaign.
“Without transparency, data partners like us can’t see what occurred during a previous campaign or the results it produced,” said Michelle Harness, division vice president of digital agency and platform sales at Wiland. “If you don’t have a view into the data and audiences used in previous campaigns, it is more difficult to make recommendations. The opacity of walled gardens makes things even more challenging.
“On the other hand, when we’re able to ingest past campaign data and perform an analysis, we can use our vast data resources to provide marketers with superior audiences and make other recommendations for their next campaigns,” she said.
This approach is especially valuable when agencies and brands share properly permissioned first-party data, including information about their top customers, with their data partners for audience selection or creation.
“The more data we can get, the better chance we can identify the most desirable customers and match them back,” Goodpasture said. “If a brand’s top customer has six email addresses, give us all six so we can find that person in our identity graph. We can then model on that data to build audiences that produce great results.”
Not all campaigns will produce relevant insights for the next one, but the cumulative effect of analyzing multiple campaigns is valuable. The most successful media teams identify and compare campaigns with similar creative and offers to see how consumers responded across channels. This often means looking beyond obvious points of comparison.
“Marketers always look at core benchmarks to evaluate performance, but there are other factors to consider,” said Kris Mann, vice president of digital agency and platform sales at Wiland. “Seasonality and other forms of cyclic behavior can be important variables to include in planning decisions.”
Historical performance data illuminates consumer response trends and helps advertisers improve targeting, messaging and overall campaign effectiveness. The KPIs for gauging historical performance will vary based on channel and campaign goals but may include clicks, site visits, video completion rates and newsletter sign-ups.
Analyzing the interaction between channels and audiences unlocks insights for future campaigns. This takes extra time that marketers might feel they can’t afford; however, in reality, they can’t afford not to make the investment. Analyzing response and conversion rates based on variables like audience, offer and creative can help to inform future campaigns more effectively.
Sometimes, this evaluation means identifying things that didn’t work, not just those campaigns that were effective. “Being transparent about the reasons for a campaign’s success, or lack of success, is the key to continuous improvement,” said Mann.
Audience selection and testing deserve substantial attention both before and after a campaign launch. It’s essential for marketers to carefully evaluate their options before choosing audience sources and segments.
Walled gardens often push advertisers to build audiences exclusively from within their toolsets. Similarly, DSPs may have dozens or even hundreds of off-the-shelf audiences meant to fit advertisers’ campaign needs.
However, the fastest or most straightforward path to audience selection often isn’t the most effective solution. Advertisers should consider all their audience options when moving forward with a DSP or social platform campaign, including building a custom audience with a data partner.
To determine which audiences will be the most effective, teams should start by making small media buys using prebuilt audiences from different providers (including walled garden audiences when applicable) and custom audiences modeled from their customer files. They can test these against each other and adjust their media spend based on audience responsiveness, as well as A/B testing various offers and creative treatments to maximize response.
“If you have the bandwidth, I recommend testing multiple audiences from various sources when you start a campaign,” Goodpasture said. “Give it a few weeks or a couple of months so that you get enough data to understand audiences better.”
While testing multiple segments is foundational to identifying and understanding ideal target audiences, agencies and brands benefit greatly from the additional context provided by enhancement data. To meet this need, advertisers work with multiple sources of third-party enhancement data — usually four or more sources, according to a recent Digiday and Wiland survey.
The information obtained from these sources — including demographic information, purchase history and personal interests — creates a more robust portrait of prospects and customers.
Looking beyond demographic and psychographic data gives advertisers a deeper descriptive and predictive view of their audience. For instance, individual-level transactional data is almost always the most predictive of future spending behavior.
“Purchase data provides insight into what an individual or household is actually consuming, as opposed to what they’re merely showing interest in,” Mann said. “Many consumers enjoy researching luxury cars, but most can only afford a domestic sedan. By using spend-based enhancement data, advertisers can personalize offers that correlate with what those consumers are most likely to buy. That makes a huge difference in generating revenue.”
Successful advertisers fine-tune their efforts once a campaign is active using dynamic creative optimization and cross-channel integration. How often a given campaign is actually adjusted depends on several factors, including its goals, channels, lifecycle and KPIs.
While advertisers face pressure to justify campaign budgets against financial results as soon as possible, they must refrain from overreacting to early returns, which can adversely impact the entire campaign.
“While you should be constantly evaluating a campaign in flight, you want to give it enough time so that the response data you capture allows you to glean reliable results,” Goodpasture said. “Then, if you change something while the campaign is running, you again need to wait and re-evaluate it within a suitable time frame.”
When returns diminish or engagement wanes, teams should test and validate everything — including offer, creative and audience — to improve performance and get back on track toward hitting their KPI goals.
“We need to know from our client brands how a campaign is going,” Harness said. “They can be in the throes of running a campaign, but if they come back to us with feedback, we can adjust the audiences to drive better performance.”
As advertisers increasingly look to implement data-driven campaigns, they must tap trustworthy, reliable data and audience partners. That process starts with an evaluation of the data itself.
“When marketers are looking for data sources, they must look at the data foundation,” Mann said. “Making sure that the data is specific and true to the actions of the individuals in question helps marketers serve each person with the best possible offer, which drives their ROI.”
Working with a data partner also helps agencies and brands in ways that go beyond digital campaigns. The same dataset can serve other purposes across their organization.
“The right data partner can benefit multiple stakeholders,” Mann said. “The analytics group can leverage the data for models and analytics, the sales teams can use it for outbound sales leads, and the marketing department can incorporate the data into its media and creative decisions.”
To effectively build on their first-party data, advertisers must seek out data and audience partners with whom they are comfortable sharing proprietary information. Only under these circumstances can they join together to analyze campaign results and understand what drove performance. Qualified partners will have established privacy guidelines to build and maintain the highest level of trust with their brand and agency clients.
Advertisers can also leverage data clean rooms, data cooperatives and other collaborative environments that bring their first-party data together with other data sets. Finding a partner whose data resources add incrementality, fill in knowledge gaps, enrich analytic environments and enable better predictive modeling is critical for data-driven success.
“What you should be looking for in a data and audience provider is a true partner,” Harness said. “Not someone who is competing with you, but someone who’s driven by the success of the agency or brand so everyone can be successful.”
