Affiliate marketing has become a steady revenue source for some of the largest news organizations in the world. Outlets such as The New York Times, CNN, and USA Today all receive commissions from Amazon and other retailers by linking their articles and product reviews to retailer links for purchase.
But would the same approach work on the local level?
In May 2019, WBUR’s BizLab, a small research group embedded within the Boston public radio station, launched an experiment to see if the station could generate revenue through affiliate marketing, driven by its online audience.
The project was supported by a $50,000 grant as part of The Lenfest Institute’s Local News Business Challenge. WBUR was one of seven grantees from around the United States focusing on experimenting with new models to support local journalism.
You can read an overview of all the Business Model Challenge projects here, and sign up below to get all of the posts sent to your email inbox.
Between May and July 2019, the team published 11 product recommendation articles, or “guides,” which received more than 39,000 unique user visits. This traffic resulted in just 111 product purchases, resulting in only $203.01 in revenue for WBUR.
Even though it made less money than it expected, BizLab learned some key lessons about what it takes for an affiliate linking program to succeed.
BizLab reflected on the experiment in great detail on its website in a series of posts here:
Here are some key takeaways:
• Get newsroom buy-in: WBUR’s digital newsroom did not agree to publish articles with affiliate links on the station’s main website or promote them from its standard social accounts. BizLab and the newsroom also had extensive discussions about the terminology they could use to describe the content. The teams agreed to call the content “guides,” not articles or recommendations.
“To clearly distinguish the content from our journalism, we couldn’t call the authors contributing writers or journalists. We agreed to call them outside contributors,” said BizLab’s executive director Joan DiMicco. “So the conversation became very much like ‘How do we make it very clear to an outsider that this is not an article’.”
As a result of segregating the content from WBUR’s news content, BizLab had to build a new website, WBUR Guides, and use costly paid Facebook ads to build an audience for the posts. BizLab also ran house ads on WBUR’s main news site to promote WBUR Guides. 850,000 users saw the promotions and 4.6% clicked through to see more content. Of those who clicked, those coming from tiled banner ads on WBUR’s site had a 0.36% product click-through rate compared to the 0.27% from Facebook ads.
• Start with Amazon: For better or worse, Amazon is the world’s largest online retailer. As a result, it makes sense to start an affiliate program where most of the readers shop. 77% of WBUR Guides purchases came from Amazon.
Along with Amazon, the Bizlab team initially experimented with the CJ Affiliate program. CJ acts as a broker for advertisers and individual brands or e-commerce sites, matching them up and facilitating affiliate marketing efforts. A benefit of using a platform like CJ Affiliate is that articles can link to many different retailers, which provides readers with a choice of where to purchase. A lot of direct feedback from the WBUR community negatively pointed at Amazon being the sole e-commerce destination for links.
Along with feedback on the affiliate marketing tools themselves, BizLab chose topics based on its audience’s lifestyle and trends. WBUR Guides featured articles on recommended fresh vegetable delivery companies, comfortable shoes, and plastic straw alternatives.
“These were all topics we thought would appeal to our engaged, socially responsible audience,” said BizLab Business Technology & Analytics Lead Ted Fuller.
BizLab suggested that CJ Affiliate should be used as a project progresses with a larger audience because its offers a higher commission, but Amazon is the best tool for quickly finding products that can build an audience. Amazon also gives publishers credit for any purchase made within 24 hours of a user clicking on the affiliate link. The variety on Amazon allows publishers to quickly test what their audience is interested in before moving to CJ for higher commissions.
• Track and measure: To find their products, the BizLab team explored many avenues, from women’s shoes to toys. Tracking audience response helped them narrow their search and tailor what was published to the highest value items.
WBUR used a variety of tools to track their audience responses. Google Analytics was installed to track overall site usage and Google Tag Manager worked in tandem with Google Analytics to specifically track actions like product clicks. While testing different campaign strategies, Google URL Builder was used to track which link distribution method brought in the most traffic. These three provided valuable user to product click data that Facebook alone could not produce.
Outside of tracking tools, BizLab sent follow up emails and embedded a survey on its site to get direct feedback throughout the experiment. New ideas as well as positive responses continued to push the affiliate marketing strategy forward.
• Focus on organic traffic: This wasn’t the first time WBUR tested out these revenue methods. The station initially explored affiliate marketing with its coverage of books. In summer 2019 — while BizLab was running its experiment with WBUR Guides — WBUR’s nationally syndicated show On Point published a Summer Book List, which included affiliate links. The station made $1,950 from 1,400 purchases.
Much of the audience On Point reached was organic. It promoted the story on air to its national audience of 3 million listeners. Only after the initial traffic diminished, did BizLab run some Facebook ads to continue to promote the page.
While the most successful Facebook ads had a high overall guide click through rate of 8.23%, mid-page banners on WBUR.org garnered, by a wide margin, the highest click through product click rate at 15.23%. Comparatively, those more successful Facebook ads only had a 3.3% product click through rate, pointing to a drop off between audience attraction and retention for actual product exploration.
WBUR had previously also used its book coverage to generate revenue, and that was actually the origin for BizLab’s experiment.
Because BizLab had to create a separate WBUR Guides website, it struggled to generate organic traffic to a brand new site.
“We were interested in finding which of the various placements were most effective at driving traffic to our site and also which audience was most likely to click on a product once they arrived there,” Fuller said.
BizLab’s prime method for promotion was through paid Facebook ads and native ad space on WBUR’s main website. Overall, Facebook ads got more immediate impressions while the native ads provided visitors who were more likely to click on the products.
This divide between initial views and actual product engagement pushed BizLab to break down the data to see this difference in action for each topic. Some topics got more immediate attention on social media, but the more niche topics that attracted the smaller, more effective audiences proved to be more valuable.
Tapping into these naturally-occurring users might bring in smaller numbers compared to those drawn in from Facebook ads, but their engagement will be much more valuable to affiliate marketing success. It would take a lot to get Facebook acquisition levels down low enough to actually turn a profit on those site visitors.
“A paid audience is not likely to be profitable given that affiliate marketing involves thin margins,” Fuller concluded, “and a paid audience is more expensive and less engaged than your site’s natural audience.”