Driving 4X ROAS for CPG Wine Client on Instacart

Brief

Bread & Butter Wines uses the online grocery delivery app, Instacart, as a central tactic in their e-commerce strategy. Because of this, our account team was eager to try a new optimized bidding tool offered by the platform. Our goals were threefold: to keep the client’s strategy in stride with a rapidly evolving platform, test AI’s ability to directly impact ROI, and reduce operational lift.

Highlights

4X
ROAS


32%
YoY Sales


$3.73
Cost per Conversion

Challenge

Coegi runs an evergreen campaign on Instacart for Bread & Butter, which has consistently delivered at or above a 2x ROAS benchmark. However, achieving these results required time-consuming manual optimizations based on cost-per-click metrics. While this approach was effective in driving results, we sought a more efficient and profitable bidding process.

Solution

Instacart’s new optimized bidding tool uses AI to automate bidding and maximize ad spend. Staying up-to-date with innovative platform updates is a priority for Coegi, and we knew this tool had the potential to significantly increase campaign performance and efficiency. Within a month of the Instacart release, our team implemented the new capability.

After a brief learning period, the AI algorithm began pushing ROAS into the 3x range. As the campaign progressed, this figure steadily increased to an average of over 4x, with peaks for individual products hitting up to 10x ROAS. As a result, the campaign generated a 32% YoY increase in Instacart revenue and decreased the average cost-per-conversion by 53%. 

In the age of Web3 and AI advances, a seemingly small platform update can have a significant impact on your results. Our team’s enthusiasm for testing and learning allowed this campaign to double its impact on Bread & Butter Wines’ ROI. 

Why the Performance Scoring Model is the Future of Marketing Measurement

 

Is your marketing measurement strategy founded in business intelligence or in media metrics? 

No single marketing metric can equate to business success. Likewise, no single marketing measurement strategy can translate success for all brands. You need a custom solution to accurately track and measure holistic brand health – based on your unique definition of success. 

This is why we believe every brand needs a performance scoring model. 

What is a performance scoring model?

A performance scoring model uses multiple, weighted data sources to define your media’s impact on business goals. This model should combine media data, business data, and advanced measurement studies, weighting each of the data points per their significance. 

Then, you can use this custom formula to create an overall brand performance score. By standardizing reporting and insights from both the granular campaign level to a broader business strategy perspective, this will allow you to make smarter, and more results-based marketing decisions.

Here is a simple example of how this formula can look:

Lift in Unaided Brand Awareness (45%) + Location Visits (20%) + Clicks (10%) + Sales (25%) = Performance Score

Using the Performance Scoring Model to Measure True Marketing Success

Advertising needs to be held more closely accountable to business outcomes. Marketing leaders are feeling this pressure more intensely now than ever. It’s uncomfortable and challenging – but these are necessary growing pains. As the industry navigates increasing consumer data privacy regulations, marketing plans require more complex planning and measurement. 

Simply put – today’s business challenges require more than basic in-platform forecasting and metrics. Media data – impressions, reach, cost-per-click – are too in the weeds to illuminate the full landscape. A performance scoring model incorporates both media and non-media data enabling marketers to make smart business decisions and more accurate predictions. 

It is simply a living, breathing business dashboard that allows marketers to accomplish three key things: 

  1. Unify disparate data sets to better contextualize and assess data analytics
  2. Clearly communicate the impact of marketing on business outcomes 
  3. Predict and inform smart campaign optimizations and strategic decision-making

3 Key Benefits of the Performance Scoring Model

1. Unify disparate marketing data sets

Data aggregation is at the core of this marketing measurement strategy. You may already be using measurement tools to combine media channels in one dashboard. But, business challenges require taking that a step further to reveal brand insights. 

The scoring model gives you a new understanding of marketing performance across the business using both conventional, and unconventional, metrics. This levels up your data analysis to go beyond engagement rates or a cost per action. You can add context by bringing in factors such as economic indicators, health trends, or any other data points impacting the business or consumer behavior. 

It’s not necessarily a tool to drive new sales or leads. But, it does allow you to frame conversations about multiple KPIs in a concise, digestible way. It can guide your marketing strategy so the media can perform better, which will impact long-term growth of bottom line metrics. Ultimately, it resets expectations and aligns teams on the incremental impact of media on business decisions. 

“With the custom scoring model, we work to see a holistic view of performance, setting meaningful KPIs and holding media accountable to business goals.”

– Ryan Green, VP of Marketing & Innovation, Coegi

2. Clearly communicate marketing results

The custom scorecard offers a more objective, quantifiable number you can use to communicate to key stakeholders. Communicating media’s value to non-marketers can be challenging at best, especially if you’re speaking with acronyms that do not apply to their daily jobs. By standardizing disparate data sets, you will be able to more easily achieve buy ins. 

For example, which of these is easier to understand? 

  • In March, FB CPMs decreased by 9.5%, CPLPV rose by 33.4%, and CTR was 1.7%. 

OR

  • In March, our overall media score was 7.5 out of 10, a 1.2 point increase from February.

Ultimately, the custom performance scorecard is a more tangible way to showcase directional return on marketing investment, in particular for stakeholders that aren’t in the marketing department (like finance or operations). Plus, it’s a very flexible data model. You can easily change the weights of each factor in your scorecard formula to accommodate input from other stakeholders or changing business needs. (We’ll get to how to create your custom formula in the next section.)

3. Inform smart marketing campaign optimizations

Finally, you can leverage custom performance scoring models to evaluate and identify leading indicators of success. You can use it to identify which parts of your media strategy are working in near real-time, rather than waiting months for results. Depending on the non-media data you incorporate, it can also help you make real-time pivots based on external factors. 

For example, you can use this model to identify highest performing DMAs. Then, you could distribute your budget and adjust messaging in softer markets versus stronger markets. Alternatively, you can swap geographic region as the optimization point with different audience groups. You can break down audiences to understand the strengths and weaknesses of each segment. Then again, strategically decide whether you will double down on strong audiences or focus on weaker audiences. 

How to Create Your Brand’s Performance Scoring Model

Ready to create your own scorecard? As you begin, media metrics are your most readily available and straightforward data points, so it’s fine if they make up the majority of your scorecard (at least initially). However, it’s important to pull in some external data as you iterate on your model over time. Otherwise, you’re siloing your marketing from other business factors. It’s like driving while wearing blinders. 

Outside perspective from non-media data guides smarter media decisions. Having that additional context can help you determine optimal frequencies, efficiencies, and top-line analytics goals. 

Examples of Non-Media Data Sources for Your Scoring Model:

  • Sales data: Sales by product/service, retailer, region, etc. 
  • Financial data: Consumer price index, stock market, interest rates, 
  • Infection rate data 
  • Net promoter score (understand your greatest customer advocates from  customers who need greater nurturing)
  • Consumer survey data: brand reputation, store cleanliness, product quality, service quality, brand loyalty
  • Advanced measurement data: sales lift, brand awareness lift, foot traffic lift

And this is just scratching the surface. You can get creative here and pull in more obscure data as long as it’s relevant to the success of your business and able to be analyzed at statistical signficance. 

Weighting Your Performance Scorecard Formula 

How do you determine what weight to give each input? I recommend leading with your intuition. But it should also be a group effort. Collaborate with the people closest to the data as well as the people closest to the brand. To avoid biases, be sure to gather input from several stakeholders:

  • CMO/Marketing Manager – Lead the discussion based on existing knowledge and marketing KPIs.
  • Data Analysts – Help provide guidance as to what data is available for use.
  • CEO/Board of Directors – Ensure strategy aligns with overarching business goals and external stakeholder needs.

As you have these discussions, remember it is an iterative process. The first formula you create certainly will not be the last. That’s the beauty of this custom model. It is adaptable, flexible, and increases in accuracy and relevancy over time as your data collection grows and your formula improves.

Implementing a Performance Scoring Model: Marketing Use Cases

Here are three ways Coegi has applied the performance scoring model to our clients:

Use Case #1 – Attributing CPG Sales to Advertising in Real-Time

Point-of-sale data lets consumer packaged goods brands see exactly how much was sold. However, the problem is speed. You often find out results weeks after a campaign. This is far from the real-time results you need to make agile marketing decisions. 

To identify CPG marketing ROI, brands typically need to go back and attempt to attribute that sales lift. Was it from your media spend? The media people certainly think so. Or did the economic boom really do all the work? Maybe it was the in-store displays… The custom scorecard model measures all of those things at once giving you a better idea of what drove sales. 

If you locate those leading indicators of success, you can have an idea of what’s working in real time. Then, when the sales data rolls in 4-12 weeks later, you can confirm what you assume to be true and adjust as necessary. 

Use Case #2 – Identifying Audience Likelihood to Travel 

The travel and tourism industry is impacted heavily by macro-environmental factors. How is the weather? What are flight tickets and gas prices? Is there a health pandemic halting travel? These kinds of factors influence where media should be placed for maximum results. 

This was especially prevalent during the COVID-19 pandemic. We used the scorecard approach for a state tourism client to create a “COVID-19 Scoring Model”. This scorecard gave each county in the state a score indicating level of opportunity for travel in each market. Using it, we were able to inform media decisions and ensure the strategy aligned with public safety. You can read the full case study here for more details. 

Use Case #3 – Identifying Highest Opportunity Geographic Markets for QSR Chain

Quick service restaurants operate in a competitive, cluttered space. Customer loyalty and share of wallet are major factors driving long-term QSR success. 

Knowing this, we create a performance scoring model for a QSR client factoring in brand lift attributes, visitation, and point of sale data. We even included data on how highly customers rated their french fries. Using this model, we were able to allocate budget to top markets and tailor messaging to boost market share among loyal customers. Read the full case study here

There are infinite ways to apply this methodology across any industry and any brand. At the end of the day, the performance scoring model is about getting to the WHY to inform the what – making our marketing strategies stronger and our clients even happier.

For help applying this approach to your brand, contact Coegi today for a discovery call

Increasing Brand Lift and Growing Market Share for BODYARMOR

Brief

BODYARMOR was a new entrant into a well-defined CPG category: sports energy drinks.  

With a product containing less than half the sugar in Gatorade, but only 2% of overall category market share, BODYARMOR was looking to disrupt the paradigm. 

 

Highlights

16%
Lift in Brand Awareness


25M
Video Views in 2 Months


23%
Lift in Purchase Intent

Challenge

The media challenge was to break through the clutter in a crowded space and ensure BODYARMOR’s message of superior hydration was reaching the most relevant audience – ultimately increasing awareness and market share.

Logistically, they secured significant investment with grocery store and gas station distribution networks. The brand also had endorsements across all major sports leagues, plus significant involvement from investor, Kobe Bryant. But, to increase market share and sales, they needed to establish brand awareness with the right customers.

Solution

We used industry research to select the optimal digital media channels, as well as our own planning and channel mix software, to develop the optimal “go-to-market” plan. Using BODYARMOR’s first-party data collected from web engagements and promotional eblast sign-ups, we created targeted media plans for key niche audiences and engaged media partners to further invest in BODYARMOR’s success and growth.

Our team performed look-alike modeling and statistical analysis to create microtargeted audiences, including: Blue Collar Workers, Grocery Gatekeepers, Veterans, Teenage Athletes, and Health-Focused Adults.

Each audience had its own media plan and messaging strategy, for example:

  • The Blue Collar Worker audience focused on Midwestern and Southern states, Facebook and YouTube channels – using creative highlighting their partnership with NASCAR’s Ryan Blaney and UFC promotions. 
  • The Teenage Athlete audience focused on Instagram and Snapchat and leveraged endorsements from the NBA’s James Harden and NFL’s Richard Sherman.

To maximize campaign efficacy, we enabled multiple layers of targeting. This included layering our media with the national distribution footprint, to ensure the campaign was reaching the right people at the right time in the right place. We customized sequential messaging based on customer engagement level and continually made real-time adjustments to optimize performance.

We also engaged our Google reps to ensure alignment and efficiency across the board. These partnerships allowed access to Google Beta products, as well as brand lift and purchase intent studies to evaluate campaign success.

By all measures, this campaign delivered superior performance. Aggressive optimization throughout the campaign resulted in 55% over-delivery and engagement rates much higher than initially outlined. Through Google Site Link extensions, we were able to drive and measure a significant amount of in-store sales volume. 

In year one, the audience-first strategy produced strong brand recall lift amongst four of the five target groups. In year two, the strategy shifted to only include the top four performing audience groups.

  • The Blue Collar audience saw 22% brand lift and $840k in attributable sales.
  • The Grocery Gatekeeper audience saw 14% brand lift and $376k in attributable sales.
  • The Veteran audience saw 12% brand lift and $411k in attributable sales.
  • The Teenage Athlete audience saw 14% brand lift, $154k in attributable sales.
  • The Health Conscious Adult audience saw 2% brand lift, $214k in attributable sales.

More importantly, sales increased nearly 300% YoY, with 6% category market share; leading to the brand being acquired by Coca-Cola.

The Countdown to Zero-Party Data

The Countdown to Zero-Party Data

If you’ve been paying attention to marketing news lately, you have no doubt seen the terms first-party, second-party, third-party, and zero-party data. These terms are critical in almost every targeting strategy conversation. 

With Google’s impending deprecation of third-party cookies, it is vital that you understand the differences between these data types. In this blog, you’ll learn how they can help or hurt your advertising strategies. Plus, we’ll outline how to collect and leverage each data source from third to zero-party data. 

Third-Party Data

What is Third-Party Data?

Third-party data is any information collected on consumers from an entity with no relationship to that consumer. In marketing, data aggregators commonly gather data from web browsers that are bundled and sold to advertisers. 

How to Collect Third-Party Data:

  • To collect third-party data, marketers purchase curated data packages from aggregators. This data is the primary target of data and privacy protection laws because it is usually collected and shared without the explicit consent of consumers. 

How to Use Third-Party Data:

  • How do you use this data? The short answer: due to changes in privacy laws/policies and the cookieless future, you should use third-party data sparingly. Additionally, this data collection can be inaccurate and lead to budget waste by serving ads to the wrong audiences.
  • Start shifting toward collecting more effective forms of consumer data, like first-, second-, and zero-party data, for your targeting needs.  

Second-Party Data

What is Second-Party Data?

Second-party data is consumer information collected directly by another organization that your brand has purchased or gained access to through partnerships. Unlike third-party data, the collecting organization has a direct relationship with the consumer. This leads to more accurate and actionable information. 

How to Collect Second-Party Data:

  • One of the more common forms of second-party data collection is through walled gardens, such as social media and retail media platforms. For example, social media account users on each platform are required to login prior to the use of the app. You may also gain access to this type of data through quality publisher partnerships. 
  • Like third-party data, you have to purchase or negotiate access to this data from the collection source. Coegi partners with providers like OwnerIQ, US Farm Data, and other reputable sources to ensure our clients have access to quality second-party data. 

How to Use Second-Party Data:

  • As mentioned above, you will likely use this kind of data while running ads on walled garden platforms or when activating direct partnerships with publishers. If you partner with a company to access this data, they will typically send anonymized email lists or require you to serve ads through them to gain access to their audiences. 
  • It is important to thoroughly vet any partnership in this space. Be sure you are in accordance with any privacy laws or policies put in place. 

First-Party Data

What is First-Party Data?

First-party data is information your brand collects directly from your audience. If you analyze it effectively, this will be one of the most important elements for digital advertising strategies in a cookieless future. 

How to Collect First-Party Data:

  • Place gated content on your website to collect emails and other information.
  • Generate email newsletter sign-ups in exchange for discount codes or special offers.
  • Store relevant information from customer purchases in your CRM platform for future segmentation and activation.

How to Use First-Party Data:

  • After collecting first-party data, you can use it to reach individuals who have already engaged in your brand through features like email-match targeting.
  • Develop modeled audiences to target people who have similar data points or behaviors to your existing customer base. Personalize advertising messages and other communications based on the most valuable and influential data points.

Zero-Party Data

What is Zero-Party Data? 

Coined by Forrester, zero-party data is collected when “a customer intentionally and proactively shares with a brand. It can include preference center data, purchase intentions, personal context, and how the individual wants the brand to recognize [them].” This data is technically a subcategory of first-party data. It is, however, worth giving this information its own terminology because it has the potential to go beyond first-party data snapshots and provide advanced profiles of your customer base. 

How to Collect Zero-Party Data:

  • Design and distribute short strategic surveys, quizzes, and polls for your audiences. 
  • Include interactive tools on your website that allow users to self-identify for a more personalized website experience.
  • Require free or subscription-based website account set-ups and logins to view the most valuable content to create a value exchange. 
  • Build product or service ratings into your website listings.

Tip: Motivate the customer by offering them something of value from your brand. For example, you could offer a discount or special access to an event in exchange for providing the data. 

How to Use Zero-Party Data:

  • Add zero-party data to your CRM and use it to curate customized communications and offers that build brand loyalty.
  • Act on user feedback to align your marketing strategy and customer touchpoints with the desires of your target audience.
  • Deliver custom suggestions to your users’ account home page based on information collected in their account set-up. 

Tip: Be conscientious about how often you are asking for this information and be sure to include variety between each ask. You don’t want to fatigue the customer and create a bad user experience. 

Bringing it All Together

The key distinction to make between each data type is the source.  As you move from third to zero-party data, you move closer to a more accurate understanding of your audience. These direct-from-the-source insights will help you make smarter strategy decisions and more effectively motivate your audience to convert. 

To learn more about how to use this data, read our Cookieless Targeting and Identity Solutions blog by Coegi’s Director of Innovation, Savannah Westbrock. 

Cookieless Attribution and Measurement Solutions

Cookieless Attribution and Measurement Solutions

Cookies have been the underpinning for most digital marketing performance measurement for over twenty years, which has allowed advertisers to measure post-click conversions and attribution for sales impact. As a result, channels like paid search and display retargeting typically stand out as ‘performance channels’. Simply put, cookie deprecation takes away the easy button of using off-the-self audiences and straightforward conversion tracking.  However, without third-party pixels, determining clear return on ad spend will become more challenging, especially for marketers who continue to rely on click-based attribution models.

Without cookies, it is imperative that you develop more meaningful ways of understanding how customers make decisions and how it impacts business results, a topic we recently covered on The Loop Marketing Podcast.

How to calculate marketing ROI in the cookieless future

In this new paradigm, marketers will need to rely more heavily on strategy to get the greatest and most accurate ROI

The ability to calculate marketing ROI starts with having a strong measurement strategy in place prior to campaign launch. Smart marketers know to look beyond online conversion data and search for correlations with business performance to determine true directional success. Advertising campaigns need to be set-up to achieve business goals rather than just vanity metrics. It’s important to know when to incorporate more robust analytical solutions to understand what’s impacting your bottom line. 

Cookieless measurement solutions

Some methods for measuring media campaigns in the cookieless future include: 

  • Media mix modeling (MMM): MMM works by isolating one variable at a time to see the impact of removing or adding a tactic. It allows deeper understanding of how omnichannel campaigns work together and incrementally impact key outcomes. 
  • Advanced measurement studies: Exposed vs. control consumer studies track brand lift, sales lift or foot traffic lift to provide greater insights into the real impact of advertising on difficult-to-measure business goals. 
  • Overlaying multiple data sources: Brands can match up Google Analytics conversion data, or sales data, with paid media data. While more time and knowledge intensive in terms of the analysis needed, this is effective to look beyond media data alone and instead looking holistically at the brand to understand marketing’s impact. 

Place less emphasis on media efficiency metrics and more emphasis on effectiveness. Look at correlations between business and media data to identify incremental conversions compared to your company baseline. 

To achieve this, marketers will need to identify leading indicators of success by channel and tactic and optimize towards those metrics.  

Will the cookieless future impact walled gardens?

Walled gardens, such as Facebook and Amazon, leverage their own first-party user data. As a result, cookie deprecation will affect them less in terms of targeting. 

Within platform confines, advertisers will still be able to track individual users, though the windows of attribution can vary. Due to this, walled gardens allow for brands to conduct some closed-loop measurement. That being said, there will be limitations on attribution, and less deterministic targeting as privacy laws continue to become stricter.

Walled garden pixels will have limited ability to pass back data to the platform once cookies are gone. We can expect front end marketing performance metrics to decline, even if backend business performance remains the same. Plan for shifts in attribution, using strategies like those laid out above, as we get closer to cookie deprecation.

Cookieless attribution tips

Begin testing and learning today to proactively understand what will and will not be effective in the cookieless future. 

  1. Begin benchmarking current performance ASAP: compare performance of cookie-based vs. cookieless tactics. Then, analyze backend data to determine the effect on business results and set expectations accordingly.
  2. Consolidate to fewer platforms, or find a way to ID map: Platforms are developing their own internal ID tracking frameworks. The more platforms you execute your media through, the more disparate measurement systems you have to consider. This will also minimize duplication across platforms. 

The deprecation of third-party cookies will undoubtedly impact the way marketers approach digital media. But a data-driven media plan tied into a holistic cookieless attribution and measurement solution will ensure your business continues to grow by reaching the audience in the right place at the right time.

3 Ways to Improve Marketing Campaigns Using AI

Are you using artificial intelligence for your marketing? 

If not, you’re likely spending unnecessary time and effort launching and optimizing advertising campaigns.

Which creative assets are best for this audience? How much should I bid for a particular ad placement? Who is my target audience? 

AI can help you answer all of these questions, minimizing guesswork and assumptions. 

How does AI boost marketing campaign performance?

To distill it down, AI uses algorithms to sort through data using a set of rules to complete automated tasks. To function optimally, the algorithm needs a goal to optimize towards. So it’s up to humans to tell the machine what that goal is and if the campaign is succeeding or not. These guardrails are necessary to ensure actions aren’t taken that create cost reductions, but are actually detrimental to marketing goals. 

With this strategic foundation in place, artificial intelligence can significantly improve campaign performance, save time, and increase efficiency.

Here are the top three ways you can use AI to improve marketing campaigns:

1) Campaign Optimization

The most prolific use for AI in marketing campaigns is for media buying automation and optimization. 

Automatic bidding allows media buyers to place appropriate bids for each ad placement in the open market. This ensures you are reaching your target spend levels and pacing evenly across campaign durations. 

Artificial intelligence tools can also optimize pre and post bid to improve campaign performance and learnings: 

  • Pre-Bid Optimization: Analyze site visitation patterns and social media following before a campaign launches for faster learnings and more accurate targeting.
  • Post-Bid Optimization: Pull insights from millions of data points to understand what worked and what didn’t. These learnings provide a roadmap for improving future campaigns and identify tweaks throughout the campaign. 

These automations allow you to focus on meaningful data and strategic thinking. Use them to repurpose your time to more strategic tasks while platforms manage the day to day operations. Look for anomalies and trends in the data, but let the AI do the heavy lifting. Then you can take a more inquisitive approach to determining the “why” behind the data points.

2) Audience Targeting 

The second way you can use AI to improve marketing campaigns is by expanding and refining audience targeting. Using AI, you are able to expand your first party and/or pixel data to find more people who look like your existing customers. Whether it be similar media consumption behaviors, demographics, or other behavioral attributes, this technology can find people most likely to convert. 

In a post-cookie world, AI-assisted targeting will be particularly important. Many brands will have less complete customer profiles due to less availability of consumer data. There will be an increasing need for AI to rapidly test various targeting tactics to find your best audiences in a way that’s both measurable and efficient. 

3) Creative Optimization

Finally, creative optimization is another impactful way to leverage artificial intelligence. Consumers are demanding more personalized, exciting moments from brands. To drive action, ads need to be hyper relevant to your niche audiences’ motivations, but also break through the clutter of other “personalized” ads. 

Dynamic creative optimization is an AI feature that delivers the optimal combinations of creative imagery and headlines to individuals in real-time. This takes the guesswork out of the creative process while improving ad personalization. By letting computers sit on top of the data, you can gain valuable insights into what creatives drive the most impact among core audiences. 

Here are a few tips to make the process of exploring AI go a little smoother:

Bonus AI Marketing Tips

  • Build a strategic foundation: AI saves time and resources, but there’s no point in capturing data if you don’t know your key business objectives.
  • Test and learn: The digital world changes fast. Use data in real-time to make decisions and see if things are working – and if they’re not, to make smart pivots.  
  • Remember the human element: We can’t forget feelings, relationships and human interaction. Make sure human insights from your team and your audience are still part of planning and optimization processes. 

If you’re a marketer, start getting more familiar with AI. Don’t expect to get it all right at first. Ask questions and talk with experts in this space. Start learning and move forward from there.

To work with Coegi and leverage our Data & Technology experts for your brand, contact us today

Cookieless Targeting and Identity Solutions

An audience-first approach or 1:1 marketing is something brands often strive for. As a digital marketing partner, it’s at the core of our mission. 

However, the ‘cookieless world’, the meanest curveball Google has thrown at the industry yet, is approaching – even if its arrival has been further delayed. With cookieless targeting, being ‘audience-first’ takes on a new definition. 

Targeting will no longer be as simple as building an audience persona and pressing “go” on pre-made data sets. Instead, it’s about really diving into the ethos of who your core consumer is and using that intel to guide your audience strategy.

We sat down with Coegi’s Account Strategy Director, Savannah Westbrock, to get her perspective and tips on how she’s helping clients prepare for cookieless targeting. The following article is an edited transcript of that interview.

It’s Time to Improve Your Audience Research

How should audience research change in light of the cookieless future?

There are three changes in audience research most marketers need to make to ensure the data tells an accurate story: 

  1. Understand the methodology: We rely on research every day to inform our media plans and marketing decision making. However, we often don’t peel back the curtain to understand how that data was collected and consider potential biases. In the cookieless future, it will be even more important to think critically and be selective with our data sourcing. 
  2. Exit the platform: Don’t rely solely on demand side platform information and forecasting for your planning. This data will be most affected by cookie deprecation. Instead, combine platform insights with external research that never relied on cookies. 
  3. Diversify your data sources: It’s time to get creative. Platform data and syndicated research will still hold value. But, you’ll need to layer it with non-syndicated data and first party data. Combine these tools to see a full picture. Even consider non-media data, such as macro-environmental trends, which may impact your audience’s behaviors and the industry at large. 

What types of cookieless data should brands be gathering to understand their audiences?

Pixel-based retargeting is essentially out of the picture. The best pivot brands can make is mining their own first-party data. But you don’t have to rely solely on your own data. Combine ‘hard’ data such as your website and platform analytics with ‘soft’ data such as social listening. Taking a more journalistic approach with these softer data sources can actually provide more meaningful insights and make your brand more authentic and trustworthy. 

Tip: Balance quantitative and qualitative data. Trust your instincts and use research to back up or refute as needed. 

How can marketers collect and expand their first-party data? 

First, you need to have systems in place to generate leads. Then, it’s all about what you do with that customer data to maximize results and become more strategic. 

Lead generation campaigns: Keep first-party data and zero-party data collection top of mind when planning campaigns. For example, promoting a useful downloadable with a lead form. This will help drive consideration and give you an opportunity to learn about your audience in exchange for shared value. 

Data enrichment: Once you collect and understand your first-party data, you can upload it to enrichment tools, such as consumer survey platforms. This helps you learn more about your audience’s interests, media consumption and day-to-day behaviors. 

Cookieless Audience Targeting Alternatives

Is contextual targeting an effective cookieless targeting strategy? 

If your audience research is thorough, you will know the channels your audience frequents, their preferred devices, favorite shows, and where they are most engaged. Pair this insight with contextual placements that make sense for your ads. 

Contextual strategies fell by the wayside in the late 2010s. Many brands focused on only reaching the “perfect” deterministic, addressable audience with cookie-based data. So some marketers may fear for impression waste by comparison. However, there are now many sophisticated contextual solutions that allow for hyper-customization and reach niche interest groups

For instance, Natural Language Processing (NLP) algorithms are beginning to better understand the actual context of ad placements using artificial intelligence. This allows marketers to implement positive sentiment targeting and smarter keyword targeting. Smart contextual offerings can optimize to real-time content trends, going beyond standard display. 

Are new user identity solutions direct replacements for cookies? 

Cookieless identity solutions such as Unified ID 2.0 and Liveramp’s IdentityLink will help reach high-value segments without wasting media dollars on the wrong audiences. But, there will still be gaps. Pre-made audiences and 1:1 third party targeting will not be the same. As cookie-based information is no longer shared across the web, we’ll need to tap a few different buying strategies. I also expect walled gardens will center in on themselves more, protecting their high value audience data. 

To overcome these challenges, marketers use all the data at your disposal to understand customers better, from channel-based information, survey data, CRM analysis, Google Analytics, and more. 

Cookieless Targeting Tips

What’s your best advice to brands preparing for a cookieless future?

There’s a lot to consider, but the two simple things brands should prioritize are: 

  1. Invest in first-party data collection
  2. Start testing now 

The most important thing you can do now is establish a baseline. Then you can conduct a true study comparing your performance with and without cookies. Cover these two bases and you will be ahead of many brands. From there, you can continue to refine and adjust your research, targeting and measurement strategies as the industry evolves. 

Our team at Coegi is actively testing cookieless solutions and brainstorming innovative cookieless media plans for our clients. For more strategic insights and tips on how to prepare your digital advertising for this change, listen to our full podcast episode on cookieless targeting here

Make Smarter Marketing Decisions with Media Mix Modeling

Every brand wants to pre-optimize their campaigns for success right out of the gate, with parameters in place to quickly attribute success to specific channels, audiences, and creatives. However, as we look toward a future without the simple attribution offered by cookies, we need to get back to our statistical roots of traditional media strategies, those without the easy button involved. That’s why marketers should lean further into the value that’s offered with media mix modeling.

What Is Media Mix Modeling?

At a high level, media mix modeling is a way to define optimal budget allocation for media channels by looking at previous campaign performance. It requires analyzing sales-related data and media metrics (Coegi’s data and technology team recommends 2-3 years worth) to make predictions and strategy optimizations that will, in theory, improve future campaign performance

Is Media Mix Modeling Superior to Attribution Models?

In short, these models should be considered and analyzed alongside each other, as they both offer valuable insights from unique perspectives.

  • Attribution gives you quick, real-time information about how specific media parameters are impacting your business goals. This information is useful when making mid-flight optimizations and short-term reporting.
  • Media mix modeling zooms out to give a bird’s-eye view of how all the pieces are working together to affect long-term strategy and performance. Each model informs the other, but tells different stories. 

How Can Marketers Build Strategies from Modeling Learnings?

It is crucial to fully understand the data you’re analyzing, not just the standard media metrics from campaign reports. What are all of the factors that may have contributed to performance fluctuations?

  • Creative?
  • Messaging?
  • Audience strategy?
  • Seasonality?

Knowing the context surrounding the numbers will give you a strong foundation to build future strategies upon. Using that context as the framework, determine what story the data points are telling you. The numbers don’t lie, but they don’t always tell the whole story. By asking the right questions, and maintaining a test and learn mentality, you will ensure strategic decisions are based on multiple factors rather than just one KPI.

How Do You Know Your Media Mix Model Is Working?

A media mix model makes predictions, but it’s not a crystal ball. Just because Facebook historically performed well does not mean it will continue to do so forever. That said, it is important to develop a continuous learning agenda to design your models.

Test your assumptions based on historical performance. For example, what will happen if you increase the budget for programmatic channels? Do overall business results change? The only way to know is to strategically make the budget adjustment and measure incremental results. From there, you can make more informed decisions about your channel strategy and budget allocation. 

Priming Your Media Mix Model For Success:

  • Keep business goals at the center of your strategy 
  • Gather quality, historical data to measure actionable results 
  • Understand the contextual factors impacting your data results 
  • Consider the sales cycle when designing tests – a longer cycle needs more time between strategic adjustments
  • Share strategic details and learnings across teams. Seemingly trivial aspects to you may impact how a model is built. 
  • Data can be easily manipulated to tell an inaccurate story. Think critically and apply business acumen to make sure you have sound methodology. 

Further Reading:

How Curiosity Can Improve Your Data Analysis

Efficiently analyzing data and providing ongoing campaign optimization recommendations is critical for driving the best results for our clients. However, the complexity and rigidity of data sets can make the most meaningful insights difficult to accurately uncover. In order to elevate your analysis and provide better insights for your campaign performance, you must stay curious, using knowledge and creativity to bolster your approach. 

Moving from Data Analytics to Data Insights

Analytics and insights – they are closely related, but not the same. Analytics is a way we approach a problem that is standardized and repeatable. But what we’re really after is turning these numbers into insights. For instance, you might see that a lot of people clicked on your ad – analytics. But going further to ask ‘What does the high click-through rate mean? and deciding how are you changing campaigns based on that knowledge – that’s insight. 

In other words, analytics are repeatable, standardized processes to understand data and gain information. Insights are what we do based on that information. 

When running a campaign, there are often thousands, sometimes millions, of data points to analyze. Even when campaign goals and KPIs are clearly defined, it can be difficult to immediately identify the more subtle and complex insights hiding within the numbers. That is where curiosity and creativity come into play – we start looking for whatever stories the data can tell.  

The Benefits of Using Curiosity in Your Data Analysis

Asking thoughtful questions and keeping an open mind about what you see in the data helps identify unexpected opportunities and avoid pre-existing biases in your analysis. This allows patterns to appear in the data and provide beneficial insights for future campaigns. No matter what kind of data you are working with, there is always a learning to uncover.

For example, perhaps you have a channel that is unexpectedly driving better sales numbers than others. Instead of broadly accepting the success of the campaign, investigate why that channel did well so you can apply those learnings and build on that success. Stretch your mind to stretch your data and let it come alive for your brand.  

Avoiding Analysis Paralysis

Anyone deeply entrenched in data analysis will know it is all too easy to get buried, and even lost, trying to extract every bit of intel from the data. And while we want to glean as much insight as possible, it can be difficult to know where to stop once you start down a rabbit hole in an effort to get every morsel of information. At some point, the return on that time investment will diminish and you’ll start coming to the same conclusions, just from different angles. 

It’s critical to strike a balance, understanding when it’s valuable to take more time to really dive into the numbers, asking a variety of questions before taking action on the insight. Quarterly reviews, post-campaign reporting, planning phases – these are all times when deep analysis and questioning of the status quo is beneficial. 

Guide Curiosity with Measurement

When you start to dig, the data can seem like an endless hole if you don’t know what you’re searching for. But if you root yourself in your overarching measurement strategy, centering on your business goals and each channel’s role in helping achieve that success, it will become much easier to determine if your media is working or not. 

It is always best to spend more time upfront talking through the strategy as a whole before you begin analysis. Then it will all come into play and be very easy. The amount of time you spend should be way more upfront than it is actually building out final data visualizations and dashboards. 

Key Takeaways

Curiosity and data analysis – yes, they do go together. Applying creativity and allowing space for human intuition and insights is key to data storytelling. But to be an effective and productive data analyst, you need to balance the art and science. Keep in mind what is and isn’t reasonable. Live for the data. Explore it and search for the stories that it weaves together. But, avoid investing your valuable time on a dark rabbit hole that leads nowhere.

Data Strategy Podcast Episode

AI Optimization – The Paperclip Theory

Paperclip Maximizing: Machine Learning And The Problem Of Instrumental Convergence

What do paperclips have to do with digital marketing and machine learning? Admittedly, pretty much nothing. The ‘paperclip maximizer’ thought experiment comes from Nick Bostrom at Oxford University. In essence, it looks at the idea that if you tell a machine to optimize to a specific goal it will do so at all costs.

If you told a machine to maximize the number of paperclips it produces the machine would eventually start destroying things like computers, refrigerators, or really anything made of metal to make more paper clips once other sources of metal run out. This concept has been coined as instrumental convergence. 

Paperclip vs Pay-Per-Click

If you transform the idea of paperclips into the idea of paying per click this becomes very relevant to digital marketing. Nowadays, almost every platform touts some version of AI or machine learning to revolutionize campaign performance, which is a boon to everyone.

By releasing control to machines, media buyers can focus on more strategic tasks such as identifying deeper insights for reports and understanding clients’ goals while campaigns continuously improve themselves. They do this by finding and optimizing for what works while avoiding the things that are not driving performance for the brand. 

See this in action on Our Work page

Defining Performance

What exactly is ‘performance?’ The easy answer is whatever your KPI may be. It could be clicks, it could be video views, it could be any trackable metric. However, this is already a simplified goal. If you are running a traffic campaign, the marketing goal should not just be to “get more clicks”.

The goal should be something designed to move the needle for the business – brand affinity growth, sales lift, etc.  For instance, driving qualified users from a target audience to an advertiser’s website and increasing brand favorability is a strong goal. This is something bigger than a single metric. A KPI can be a stepping stone and an important indicator of success, but it is not the final objective.

A truly successful campaign will not simply be the campaign that drove the most clicks at the cheapest price point. Success lies in the campaigns that drive true performance towards core business objectives.

Looking for a partner to harness AI technology and drive marketing performance? Reach out to Coegi for a discovery call today.

What Machines Lack

Context. Context is something that a robot has not yet mastered.

As a marketer, I know that increasing clicks, directionally, should push me closer to my goal. The machine knows this too, but this is all the machine knows. It will endlessly optimize to a single goal.

Maximizing the number of clicks given a fixed amount of budget. This can lead to unintended consequences – think back to the paperclip example.

A machine might say only run display banners and forget about high impact formats such as video and CTV. The machine might push 100% of impressions into in-app environments. It would say never buy another out of home ad again. The machine would never know to build brand awareness, because there is no optimization point it can use. 

Machines Need Guidance

We understand the role of the media buyer is not going away, but it is morphing. There is a new symbiotic relationship between buyer and machine which empowers them to maximize your brand as a whole. Successful media buyers do not need to spend 80 hours per week finding every winning media combination.

Most of these tasks can be done through harnessing technology and freeing up time to look at the media plan from a higher level.  These benefits of using AI go directly to our clients in the form of more time dedicated to listening to client needs, smarter digital media plans, and ultimately higher performing campaigns. 

How To Incorporate Smart AI In Your Media Buying

  • Release media optimization controls to AI machines and spend time on strategic campaign elements. 
  • Define performance success beyond the metrics by establishing meaningful campaign goals 
  • Use context to avoid instrumental convergence and potentially harmful optimizations from unattended machines.

Click the button to view our full YouTube playlist on AI for Marketers.

Coegi Partners

/ Contact - usa

Tell us about your project

This field is for validation purposes and should be left unchanged.

Coegi Partners
Skip to content