Look, I’ve been at this long enough that I’ve seen it many times before. “There’s a great new way of doing things that will disrupt marketing.”
Sometimes it is truly revolutionary (the internet). Other times it’s the next great hype that never materializes (Second Life, Google+, MySpace, Google Glass, The CueCat, etc.).
Will artificial intelligence and machine learning be the former or the latter for marketers? And how can marketers best utilize it right now…and in the near future?
It’s a topic we continue to explore, so today we bring you examples from SAP, a TV network, and an adventure travel marketplace.
This article was published in the MarketingSherpa email newsletter.
“Apply to join our new research cohort where we will help you test and build an AI-calibrated MECLABS SuperFunnel,” Flint McGlaughlin offered in Landing Page Conversion: 4 powerful ways to develop a cohesive, effective strategy.
It’s part of our effort (MECLABS Institute is the parent organization of MarketingSherpa) to work with marketers and entrepreneurs to determine how to best use current and emerging AI-driven technologies to build an effective funnel.
But rest assured, to serve you, this reporter has to remain a skeptic of artificial intelligence, machine learning, and really, every other “next best thing.” In much the same way I’m guessing a magician can’t muster up the suspension of disbelief necessary to watch someone else’s magic show, after a lifetime in the marketing and advertising industry I just can’t hop on a bandwagon without a healthy dose of skepticism.
So, to help you determine when to use (and avoid) AI, today we bring you two stories of when brands leveraged AI…but also a story about a brand that avoided using the technology.
First up, a television network that used artificial intelligence and machine learning to get more traffic from organic distribution. Then, how SAP used AI to avoid cookies and increase brand awareness with a paid campaign. And finally, an adventure travel marketplace that invested in humans instead of AI to increase sales from organic traffic.
Sistema Brasileiro de Televisão (SBT) is a free-to-air Brazilian broadcaster with 114 television stations and over 6,000 employees. The network’s website attracts 11 million unique viewers and more than 99 million pageviews per month. The TV network as 12 million followers on its main Facebook page.
On an average day, SBT posts between 100 to 150 pieces of content to Facebook alone.
“We do not have a centralized social media team, so every content production team ends up posting independently,” says Rodrigo Hornhardt, Journalism Integration and Planning Manager, Sistema Brasileiro de Televisão. In total, he estimates that around 20 people throughout the company post regularly to SBT’s social media accounts, rendering the post scheduling process time-consuming and convoluted.
SBT posts content to a wide variety of social platforms, but Facebook is the most important. According to Hornhardt, “around 40% to 50% of our web traffic comes from Facebook.”
To solve these challenges, SBT turned to marketing automation technology powered by artificial intelligence (AI) and machine learning.
Before, SBT’s staff had undertaken the laborious task of posting new content manually — a workflow that was as time consuming as it was inefficient, especially given the decentralized nature of SBT’s social media management. “With so many people posting so much content across the company on different platforms, AI is so important in keeping everything aligned,” notes Hornhardt.
SBT’s reasons for selecting an AI solution were manifold: “The possibility of having AI recommend the best content to post was important, but so, too, was the ease with which multiple people can manage posting for a unified output. The ability to automate workflows was also a key consideration,” he said.
To increase the reach, visibility, and engagement on posts, the team makes regular use of real-time platform trend data alongside SBT’s own audience data to determine the best time to share content. This not only reflects audience habit — when do SBT’s readers most often engage with posts? — but also the ever-changing factors of Facebook’s Feed algorithm. Hornhardt notes that SBT News in particular uses AI to understand the optimal time to publish and maximize the impact of the multiple news items it posts each hour.
AI technology determines optimal share timing by constantly analyzing and computing the predicted performance of shares to maximize their inclusion in the Feed. Using advanced machine learning, it becomes possible to continuously reverse engineer the workings of Facebook’s algorithm, producing a sophisticated and accurate picture of the best time to post to drive engagement.
When posts are shared, AI automatically continues to update each post’s optimal time based on the latest data, pushing certain posts back if new, higher-potential posts are added, or bringing posts forward if their short-term potential goes up. This is just one example of how using AI technology to augment SBT’s social sharing has meant, in Hornhardt’s word’s, “increased reach and organic and substantial growth.”
By incorporating artificial intelligence into its social media strategy, SBT saw strong improvements in performance on social media. Within four months of adopting this AI technology, SBT’s social media pages saw increases in daily clicks of 25%, whilst daily organic impressions rose by 61%. On Facebook, SBT saw a 52% increase in daily impressions, while the company also saw improvement gains on Twitter — pageviews from the platform increased by 63%.
In addition to these performance gains, SBT has increased workflow efficiency using automation, bringing time savings in the process. Within four months of adopting AI, the team shared almost 40,000 posts; making these shares via AI and using automation has saved the company 14 hours per day.
“Brands and content producers can greatly benefit from intelligent automation, but it must be tailored to their content and audience,” says Simran Cashyap, Head of Product & Design, Echobox (SBT’s content automation provider). “While each brand must experiment to determine the ideal level of automation for them, we tend to find that the more a brand automates, the bigger the performance gains, as the AI algorithms have more opportunities to continuously learn from the content and audiences.”
By relying on AI’s capability of calculating the optimal share time, the team has increased traffic and impressions whilst saving significant amounts of time. Daniela Nobre, Customer Success Representative, Echobox, explains AI and automation “completely overhaul a company's organic marketing strategy by digging into granular data, something little to no marketing teams could try to replicate.”
For Hornhardt, this is only the start of what could become a more specialized social strategy. “We need to dedicate more working time to planning and defining strategies for networks,” he told us. “One way would be to have specialists dedicated to this strategy with the possibility of creating an audience/networking team. We see AI supporting us in this endeavor.”
The demise of the third-party cookie has left data-driven companies with a new focus: to build cookie-less alternatives into their marketing strategy. And that’s exactly what global software company SAP did this year.
The team has always had a robust marketing strategy in place, but with the end of the third-party cookies looming, it recognized the increasingly pressing need to focus on privacy-first, cookie-less campaigns.
The team chose to place its ads alongside contextually relevant content, targeting key business decision makers based on what they had chosen to look at in the moment.
They used contextual targeting technology that uses AI to see and analyze the full content of a page by scanning all of the main content data signals – from text and imagery to audio and video – for accurate and safe ad placement, without the need for third-party cookies or any personally identifiable information (PII).
The team used a wrap-style ad that wraps around the content the visitor is viewing. The ad had an entrance animation to attract the user’s attention. The right panel had a countdown to the advertised webcast. And the left panel had a message and CTA.
Creative Sample #1: Wrap ad for SAP
When the user scrolled, the background changed to another visual. The headline was repositioned onto the left panel, so it had a consistent presence.
Creative Sample #2: Scrolled state of wrap ad for SAP
“Throughout the campaign, the aim was to boost brand and product awareness,” commented Moritz Fisecker, Integrated Media Manager EMEA, SAP.
In order to successfully establish the impact of SAP’s campaign, the team analyzed the difference in results for a control sample (users who were not exposed to the high-impact ad creative) and the target audience (users who were exposed to the ad creative).
Finally, they used eye tracking to understand the level of consumer attention the ads garnered.
“Overall, the campaign results proved that brands can reach consumers just as effectively – if not more effectively – using a completely cookie-less approach to digital marketing rather than relying on behavioral data,” Fisecker said.
More than half (61%) of users exposed to the campaign took or intended to take some sort of action; and the ads had a 93.2% viewability rate.
Others results included:
In addition, the study concluded that the ad creative drove significantly more users to become “very interested” in finding out more after being exposed to the ad.
“In addition to awareness, Lumen Research and On Device Research (ODR) – SAP’s measurement partners for the campaign – confirmed that users who saw the ads had a heightened interest in finding out more. The results also affirm the role of contextual data in the future of mar tech. In this new cookie-less, privacy-first era, tapping into the customer’s mindset ‘in the moment’ will be the key to inspiring them, and ultimately, influencing their behavior,” said Peter Wallace, General Manager EMEA, GumGum (SAP’s contextual intelligence partner).
10Adventures is an online marketplace that connects adventure travel enthusiasts with local guides around the world. In addition to allowing users to book adventure travel, it recently launched a subscription-based GPS-trail app that enables people to safely explore the outdoors.
“We launched 10Advnetures in 2019, with a WordPress MVP (minimum viable product) and great content,” said Richard Campbell, Founder & CEO, 10Adventures. The site currently has more than 1.5 million annual unique visitors.
Let’s take a look at the most – and the least – effective tactic the team used to grow website traffic.
“We have had an incredible journey to build our skills in ranking high in Google,” Campbell said. “There are literally dozens of small bits of work we have done to get our content to the top of Google and Bing.”
The team invested the time necessary to create engaging, factual, useful content that targeted the right keywords.
“Users want quality content. The web is filled with superficial content, written by content farms with lots of mistakes. This is especially true in the outdoors market,” Campbell said. “In the world of content, the cost of efficiency usually comes at the expense of quality. Many businesses will turn to cheaper or faster solutions like AI writers or outsourcing to meet their goals but won’t see the desired performance if the quality isn’t there.”
The team’s most popular content is their free route guides, especially regional pages of the best hikes (or road bike rides, backpacking routes, cross-country skiing, etc.) in a geographic region.
Creative Sample #3: Route guide content on adventure travel marketplace website
The team then implemented the following elements of on-site SEO:
For external links, the team tried third-party backlink providers but found that doing it themselves was cheaper and ensured that all backlinks were organic.
The team felt a Google algorithm change punished the site for having slow site speed. “Algorithm changes are a constant reality, and Google places plenty of demands on websites to perform,” Campbell said. “Being listed first in Google is tough, and recently site speed has been an important metric that we felt was hindering our ability to be first in Google.”
Campbell continued, “Unfortunately, our site was based on WordPress, and we had done everything we could to improve site speed.” This included identifying performance bottlenecks and optimizing the databases and code behind them.
Ultimately, though, the team had to implement a new solution, and moved to a static site, rebuilding the entire front-end in a new language. This meant each page is rendered and doesn’t need database calls, creating a quicker user experience and greatly improved page speed performance. In addition, they now have edge servers around the world. In the month since this change, organic traffic is up 67% compared to last year.
For this reason, the team has baked in regular content reviews to look for opportunities to rank higher, as well as evaluate the website’s performance.
The team found that paid marketing was not only expensive when compared to what they were doing with organic, but it also resulted in a lower conversion rate – their paid advertising leads convert at 10% of their organic leads.
The team has created many permutations of different ads, but in general can’t replicate what worked pre-pandemic. In 2019 when they launched, customer acquisition cost was about $300. Since the end of last year and the first seven months of this year, it was around $1,000.
Thanks to the traffic generated by their organic tactics, they now have a registered user base of more than 50,000.
The previously mentioned increase in organic traffic has also led to growth in another key area of the business – sales. When looking at the 10Adventures marketplace and tour sales over the last 18 months, the team saw steady growth that has culminated in nearly $1 million of sales over that period, and almost 5x growth year-on-year.
Going forward, the team is working on experiments with their new site to understand their ability to influence rankings from on-page rewrites and other small-scale optimizations. They want to know if it’s easier to move content from third place to second, from eighth to third, or from 40th to eight, etc. Does this ability vary based on the type of content, or the location the content describes? These discoveries will guide their work in the coming year to continue to grow their organic traffic.
Get Better Business Results With a Skillfully Applied Customer-first Marketing Strategy
The customer-first approach of MarketingSherpa’s agency services can help you build the most effective strategy to serve customers and improve results, and then implement it across every customer touchpoint.Get More Info >
Marketer Vs Machine
Marketer Vs Machine: We need to train the marketer to train the machine.Watch Now >
Free Marketing Course
Become a Marketer-Philosopher: Create and optimize high-converting webpages (with this free online marketing course)See Course >
Project and Ideas Pitch Template
A free template to help you win approval for your proposed projects and campaignsGet the Template >
Six Quick CTA checklists
These CTA checklists are specifically designed for your team — something practical to hold up against your CTAs to help the time-pressed marketer quickly consider the customer psychology of your “asks” and how you can improve them.Get the Checklists >
Infographic: How to Create a Model of Your Customer’s Mind
You need a repeatable methodology focused on building your organization’s customer wisdom throughout your campaigns and websites. This infographic can get you started.Get the Infographic >
Infographic: 21 Psychological Elements that Power Effective Web Design
To build an effective page from scratch, you need to begin with the psychology of your customer. This infographic can get you started.Get the Infographic >
Receive the latest case studies and data on email, lead gen, and social media along with MarketingSherpa updates and promotions.