A 5 Point Plan for Quality AI Creative

Hit Short Term Goals While Maintaining Long Term Brand Favorability

‍

CMOs are under pressure to reduce costs, increase efficiency and deliver against short term goals. This pressure can create a ripple effect for everyone along the advertising supply chain - from agencies to ad-tech companies to publishers. Under pressure, people loosen standards and cut corners, which might provide a lower short term cost or time savings, but hurts the brand on the metrics that matter. A few hours saved now at the risk of lower brand favorability later is not worth the tradeoff. 

‍

One recent study from the MMA shows that brand advertising is critical for driving long term results and that companies underestimate the impact of brand on sales by up to 83%. The MMA notes: To achieve the best possible returns or outcomes from marketing and business investments, a significant portion of resources should be allocated toward building and strengthening the brand.

‍

Effective brand marketing requires an adherence to brand standards and a use of measurement metrics that go beyond immediate clicks and conversions. 

‍

AI Must Support Brand Building

‍

AI makes it easy to generate creative, which can be a slippery slope for advertisers and agencies. Just because anyone can make as much creative as they want any time doesn’t mean that creative adheres to brand standards or performs effectively against important KPIs. 

‍

Maintaining high quality standards is a must to ensure AI creative delivers results for brands. With that in mind, we’ve created a five point plan to ensure your AI creative delivers against short term goals without undermining long term brand favorability.

‍

Point 1: Involve creative experts

‍

While AI can generate incredible creatives, this capability should not exist in a vacuum - it creates risk and it eliminates the inherent value of professional creatives. 

‍

Putting AI tools in the hands of professionals reduces the risk that AI creative will be problematic. Creative generation is the most common use of AI in our industry according to the IAB, at 54%. The report notes the inherent risks in the adoption of AI without any guardrails including “misinformation and deepfakes, loss of creative control, and brand integrity risks from offensive or harmful outputs.” People need to check AI because AI doesn’t operate by the same rules that humans do and doesn’t see what people see. An AI generated creative might touch on a culturally sensitive nerve, contain an obvious error or impossibility or simply render or exaggerate images in a way that doesn’t resonate with a brand’s image. 

‍

Experts ensure that AI creative maintains inherent brand standards. Studies show that people do like AI images, but immediate results are not always indicative of long term outcomes. Consider the classic case study of the The Pepsi Challenge. By judging on a single sip of Pepsi compared to Coke, people chose Pepsi because it was sweeter. But over time, people preferred an entire can of Coke. Judging future overall performance on the first sip can be dangerous. 

‍

Experts can ensure that AI images don’t optimize too strongly for near-term KPIs, that they are sensitive to relevant context, and can use emotion, empathy and creativity to bring new ideas to the AI process. 

‍

Point 2: Increase expectations 

‍

There are many things that AI can do that were previously unattainable, which should make brands raise the bar, rather than keep their expectations where they are and simply reduce overall time and costs. The reason? Competition. If one brand starts to push creative boundaries with AI, they will quickly surpass competitors who stay where they are.

‍

If AI makes creative development easy, advertisers have an opportunity to reach new goals that were previously unattainable. Consider these examples:

‍

• Shift from preplanned product placement to AI-generated product placement that’s programmatic and scalable. Rather than simply saying “let’s save time by using AI to place a product in that scene,” brands can increase expectations to place more products in more scenes, and even dynamically render products based on audience data. 

‍

• Generate more relevant creative for each platform. Instead of trying to save time by auto-generating the same creative for each platform, AI can be used to make a variety of creative that is contextually relevant or personalized for specific placements. 

‍

• Break boundaries across creative formats. Brands can move into platforms and channels that were previously too costly like CTV, or they can experiment with combinations of creative that they previously couldn’t scale such as using social influencer content on a video ad or dynamically generating ads targeted to past shopping behavior in real time. 

‍

If the first step in AI is to reduce the manual work it takes to reformat and resize ads, the second step is to use that newly freed up time to generate better creative in more places. 

‍

Point 3: Test new innovations 

‍

Hand in hand with increasing expectations is creating an environment that rewards testing. Continuous experimentation works, and too few advertisers do it. AI unlocks completely new concepts such as in-content advertising at scale, but advertisers need to create the right incentives and processes for their teams and partners to unlock these opportunities.

If teams are required to focus on short term goals to the detriment of all other activity, there will be no room left for testing and innovation. Brands need to carve out space - both time and within team goals - to get to know AI and understand its potential.

‍

Whenever a new capability enters the advertising ecosystem, there are brands that don’t see performance improvements from a first test and become overly cautious or skeptical. A testing culture requires brands to accept that certain experiments will perform below expectations as teams learn and as processes get ironed out. For example, a first test with AI generated video might not perform as well as a traditionally produced video, but with a few tweaks such as different background music or the addition of a dynamically generated product image, the creative might become the top performer.

‍

With AI, the sky really is the limit, and advertisers that empower teams to continuously improve and optimize as they test will be light years ahead of the rest. 

‍

Point 4: Collaborate with partners 

‍

An AI program will come together more quickly when it’s founded in strong partnerships. Agencies, creative tech companies, publishers and ad-tech companies are on the front lines of working with AI generated creative. Brands can learn quickly if they stay in the loop and lean into the right partnerships.

‍

Keep in mind that until now, most advertising partners, from media agencies to ad-tech companies to media companies are used to creatives that are handed to them from another source. This is a new era, where everyone suddenly has tools that allow them to manipulate creatives - in many cases without the internal expertise or knowledge they need to focus on quality. 

‍

Ask partners the following questions to create an AI program that ensures best practices:

‍

• How are other brands working with you to ensure AI creative is high quality?
• What guardrails do you put in place to adhere to brand standards?
• What metrics do you use to ensure creatives deliver against long term KPIs not just immediate performance metrics.
• Can you explain your process and how your team works with AI technology exactly?
• If I’m focused on quality, what best practices can you recommend and how can we follow them together?
• How can I goal you as a partner to uphold quality standards over time?

‍

Brands need to be actively involved - finding partners that are creative experts with a deep understanding of AI and limiting the ability of partners who are not creative experts to make unwarranted changes. 

‍

Point 5: Lean on KPIs

‍

Advertisers should be diligent about measuring the value that each creative delivers using metrics that tie to real business outcomes, not just proxy metrics. As advertisers test new things, the goal is to fail fast, learn and improve, creating a positive loop of improvement over time. This only happens when brands are diligent about balancing short term outcomes with long term value creation.

‍

Consider these categories of metrics:

‍

• ROI and ROAS - These metrics will help capture the spread between cost and performance of a creative asset. The cost must include both creative production and media. A creative that takes more time, is tested, and includes human expertise will be more expensive than an auto-generated image, but if it performs better, it will require less media to deliver on key goals than the cheaper ad.
• Brand Favorability - During certain times of the year, brands push hard to drive sales and hit short term goals. For example, offering steep discounts over the holidays. This practice has been shown to be overused, actually hurting overall revenue and degrading brand favorability for the most loyal clients. The same concept holds true for low quality creative. Pumping out inexpensive messaging might cut short term costs, but will erode brand favorability in the long run.
• Loyalty and Churn - By focusing on the best customers, brands can see if new creative approaches are creating issues or delivering value. In a given campaign, a creative might attract a host of net-new shoppers to a website, but if they all leave, or only purchase once never to return, the creative didn’t do as good of a job as one that delivered lift across returning customers and brand loyalists. 

‍

AI Integrity Delivers Long Term Value

‍

In advertising today, every interaction matters and every creative counts. Brands compete for attention, time and share of wallet and that competition is increasing over time. 

‍

While it’s tempting to use AI for quick cost savings - ignoring the importance of quality backfires. Pumping out low quality creative to get through Q4 will create problems the next year as performance starts to go down. A focus on quality keeps creative from going sideways. 

‍

We understand the balance between AI efficiency and creative quality. When advertisers balance the two, it’s possible to reduce costs and time spent while increasing the output and opening up new advertising opportunities that were never before possible.

‍