COVER STORY: What happens when Google changes its algorithm?

The fight to the top of page one on Google results, or for the most Facebook impressions is a dangerous game. E-commerce businesses have seen both incredible successes and spectacular failures from chasing search engine and social media algorithms.

While algorithmic business models can provide massive opportunities for reaching new customers, relying too heavily on third-party algorithms can expose businesses to enormous risk, as suppliers can change their algorithms without notice, leaving these organisations scrambling.

Shane Lenton, founder of The Wishlist Company and Unified Commerce, and chief information and digital officer at Cue Clothing told Digital Nation Australia that he has seen businesses with this model “decimated overnight”.

“There's a lot of businesses out there that have put all their eggs in one basket, per se. And they've seen some success over the years, incredible success, but now overnight they've seen the rug pulled out from under them,” he said. 

According to Lenton, one particular business had as much as 90 percent reliance on Facebook for revenue and was investing up to $1 million per month in advertising, before Facebook changed its algorithm and the entire business model was “flipped on its head”.

Search engines are not designed for you

Adam Heimlich, co-founder of Chalice Custom Algorithms told Digital Nation Australia that the core problem is a misunderstanding around what the algorithms are designed to do.

“We would say the core problem is the algorithm they're using to make decisions wasn't built for them,” said Heimlich.

“The platform algorithm that's trying to predict for many advertisers at a time, isn't necessarily trying to get as many conversions for any single advertiser as they can.”

According to Heimlich in order to predict for many advertisers at a time, the difference in business outcomes is reduced to overly simplistic metrics like clicks and last touch conversions.

“[The metrics] are supposedly good enough for everyone, but they're not good enough for everyone and especially large advertisers who have a significant word of mouth, clicks could be a terrible signal of advertising value for them.”

In order to mitigate the risks of third-party algorithms, Daniel Benton, general manager at Neo Media World Australia, said that diversification is key.

“If you are just reliant on one channel, you are super exposed because if that's your primary source of traffic and primary source of revenue, it's a tricky proposition to be in,” said Benton.

“The way that brands need to think about it is to be hedging their bets, whether it's around building out more channels, so building out first-party data for your CRM program, where you can communicate with people via email, SMS, et cetera. Looking at channels like affiliates as an example to de-risk the bigger platforms.”

According to The Wishlist’s Lenton, building out a comprehensive CRM system is critical for businesses over the long term, as opposed to chasing the illusive algorithm.

“When you are onboarding those customers, make sure you’re putting them into a funnel. You've got the relevant automation, it will drive far more profit into the business and obviously improve your margins by doing so. Then using your own algorithms through your own channels, your marketing automation platforms and your CRM,” said Lenton.

“For success, from a long-term perspective, but even midterm, it's critically important to think about what you're doing once you acquire the customer and once you get that first sale.”

Building a relationship with the customer is crucial in converting them from a single purchase to a second or third purchase, he said.

“Driving that frequency, the return investment, that's where it's sustainable and ultimately going forward it's a far cheaper cost to sale," he added. 

Brand and CX

Neo Media World’s Benton believes the biggest differentiator for e-commerce businesses is building brand.

“The former CEO of Google, Eric Schmidt came out circa 2010 and said, ‘Brands are one of the factors we use to actually sort and determine who deserves to be on page one’. If Google can determine that you are an authority, for example, clothing like The Iconic or financial services like Finder, that's going to help you to show up. It's going to give you topical authority,” he said.

Benton said improving brand will drive stronger performance, better engagement and increase trust among consumers.

“The challenge is it isn't easy, and unfortunately, it's capitalism. It's based on, who's got deepest pockets and can invest the longest in the most places.”

Customer experience is also critical, in today’s environment where organisations are not benchmarked against their peers but are benchmarked against the best.

“You're getting benchmarked against say Commonwealth Bank and finance, which has got phenomenal UX or Uber, Amazon, or Airbnb. You've got to have a great digital experience. You’ve got to have a great product. You've got to have great things like customer service and delivery. All of these elements come into play. If you can get it right it can be very powerful,” he said.

Gaming the algorithm

While Chalice Custom Algorithm’s Heimlich said that there are benefits to knowing how platforms like Google work, he doesn’t see the value in chasing the algorithm.

“I started as a Google search buyer, right at the beginning around 2002. If you called Google for help and asked enough questions, they would just put an engineer on the phone that would talk to you about how the predictive algorithm worked. I had that experience, but 20 years later, Google is completely different now. You’re not going to find out anything about how it's working,” said Heimlich.

“It's very limited how much can be shared about their modelling and predicting. So, you should know what you can and you should guess as you can, but what you shouldn't do is rely on them for prospecting.”

Neo Media World’s Benton warns that attempts to game the algorithm are not just a waste of resources, but can actually be damaging to the business, as it goes against Google’s guidelines.

“It can work, but there's a risk associated in that your site can be penalised,” said Benton.

“It's in breach of the guidelines and you can be penalised, you might get a short-term sugar hit in terms of traffic and conversions, but it's just not a long-term play. Marketers need to understand if they are doing this, what is the potential downside if we get penalised?

"What is that worth? How big is this traffic source? And generally, if it's search, SEO will be between 20 to 70 percent of a site's traffic and whatever percentage of revenue, losing that can be pretty catastrophic.”

Opportunities

Despite the risks of algorithmic business models, when leveraged appropriately, third-party algorithms can be an extremely useful and beneficial tool for businesses.

Carl Bischoff, managing director of Clicksology was emphatic about leveraging search through the Digital Giants.

“It's not a problem, it's a bonus, it's a positive, it's a downright win!” he told Digital Nation Australia.

“Facebook's got all this data for heaven's sake, use it! It's like, duh! People try and go and sort of target specific things in Facebook and like holy crap, Facebook knows all about it before you even turn it on, they’ve reviewed the page.”

While Bischoff claims to “love” search and social platforms for the reach they offer, even he recognises the challenges.

Commenting on Google’s new Performance Max strategy, he said, “I don't like it because I've been running Google ads for almost 20 years. We know what we are doing and we like to be able to pick and choose what we're doing. Now it's like, ‘No, you don’t have any control, we're just going to run it for you’. So that’s frightening.”

Capitalising on the disruption

According to Lenton, the businesses that can gain the greatest advantage from search engine algorithms changing are those that can capitalise on the disruption.

“The businesses that do understand pretty quickly what the changes are and how to capitalise on them, there's a huge opportunity,” said Lenton.

“But fundamentally lessons need to be learned and the cost of acquisition of customer is so high. It's critically important to focus on the lifetime value of the customers and the lifecycle of those customers, bring them into your data and understand who they are, what they want, use that data and use the algorithms that you can control to ultimately drive repeat purchase.”