How AI in Ecommerce Enables True Personalization: Q&A; With Elizabeth Gallagher of Lineate

“It’s machine learning’s job to find patterns based on the data you give it to help you focus on the data points most likely to lead to conversion.”

Elizabeth Gallagher, chief revenue officer at Lineate talks about how machine learning (ML) and artificial intelligence (AI) are changing the game for ecommerce brands. With the use of predictive analytics, marketers can create personalized marketing campaigns.

In this edition of MarTalk Connect, Gallagher shares the key data points marketers should use to provide personalized recommendations. She stresses how data-driven automation and machine learning are strategic assets to enhance the customer journey.

Key takeaways from this Q&A; on AI and machine learning in ecommerce:

• Learn how to streamline every aspect of the customer journey
• Tips for legacy brands who are starting to explore AI and machine learning in ecommerce
• The latest AI and machine learning trends for 2020 and beyond
   

Here are the edited excerpts of our conversation with Elizabeth Gallagher of Lineate on AI in ecommerce:

With technologies such as machine learning (ML) and artificial intelligence (AI) changing the game, how can brands determine if they are providing truly personalized interactions?

It depends on the quality of your data and how you’re hoping to personalize interactions. It’s important for brands to remember machine learning algorithms need a specific output or metric to optimize for. Let’s say an event company wanted to provide music lovers with personalized recommendations for shows. They’d need to train the algorithm on users with similar tastes, ticket purchase history, as well as that user’s on-site behavior. From there, it’s quite simple: are the users for whom you’re personalizing taking more action than those who are receiving more generic marketing messaging? If not, it’s safe to say there’s more room to tweak your algorithm.

Which key data points should marketers use to provide personalized recommendations? How does AI and ML play a part in such personalization?

In order for a predictive analytics algorithm to work, you need data points that will allow you to you guessed it, predict who your customer is, what they care about, and what might influence them to buy. Now, this can stem from their own historical data like past shopping history, site browsing patterns, or carts they’ve abandoned. Brands should also pull data from customers with similar backgrounds for further clues into what nudges certain groups of people to buy. Once you’ve identified ones you’d like to predict, it’s ML’s job to find patterns based on the data you give it to help you focus on the data points most likely to lead to conversion.

Learn More: 13 Digital Marketing Trends for 2020 From 14 Top Marketers

To what degree can marketers influence customer behavior and use predictive analytics to create personalized marketing campaigns to reach a customer at the right time and right place?

There are two main ways that marketers can impact the right time/place aspect of personalization. The first is by finding patterns in the relationship between location and engagement in the historical data brands have gathered (i.e. travelers most often upgrade airfare tickets when they are already at an airport). Machine learning can then make predictions based on these patterns and marketers can plan messages accordingly (in the previous example, it may be to add an “upgrade your seat” button in the flight check-in reminder SMS).

The second is by finding patterns in correlations between when people typically make purchase decisions (deciding to buy a cappuccino or breakfast sandwich before they start a morning commute) and using that data to send right-time messages within those timeframes. (In the example would be to send a coupon to a certain market segment at 7:30 am.)

In what ways can customized technology solutions help brands retain customers longer by using content recommendation or ensure that offers are sent where and when they matter most with geo-targeting?

Most of the marketing and advertising channels provide some form of geo-targeting. Where custom technology comes into play is digging deeply into engagement patterns to make solid content recommendations. Netflix and Amazon have made this look easy, but getting it right takes some real analysis of your content and customer engagement. It also takes customization to build out recommendation algorithms that meet the needs of each brand’s content/products and customer base.

For geo-targeting to be effective in a consumer marketplace where real-time personalization is now expected, ensuring that the targeting across channels and devices and the recommendation algorithms are tightly bound is critical.

How can tools/solutions help streamline every aspect of the customer journey from acquisition, to retention, and engagement?

I think it’s commonly accepted that breaking data silos of customer data is a requirement for personalization across the customer journey. Aggregating this data together including purchase history, abandon cart email clicks, display ad engagement etc., is a prerequisite for succeeding in today’s climate. Additionally, as data regulations increase respecting customer consent across the journey is critical to a positive Customer Experience, having all your data together is necessary for that. For most marketers building this data technology is not feasible or necessary so adopting CDPs or data orchestration tools will be the most impactful change to make.

What are the 3 crucial steps that marketers need to be aware of to deliver a seamless customer experience (CX)?

1. Unify customer engagement data
2. Leverage your data to personalize your experiences across all touchpoints
3. Respect consent and best practices to instill and foster trust
    

What are your top 3 tips for legacy brands who are starting to explore AI/ML in e-commerce marketing? How can marketers integrate AI and ML from inception through to measuring performance?

1. Get informed about what ML can do for marketers and what you need to get started. Like all exciting technologies, there is a lot of noise out there and some over-inflated claims. A dose of realism will help guide you forward.
2. Start small! For example, implement AI-driven upsell recommendations based on the page of one of your more popular products rather than your entire catalog.
3. Define a clear hypothesis about which metrics indicate improvement and how much improvement you expect to see. It’s crucial to follow a scientific method to avoid getting lost in the complexities and possibilities of ML. And, automate collecting the performance data so the results are transparent and easily measurable.
4. Outsource for expertise – it is unrealistic (and time-consuming) to spin up a data science and ML group for these kinds of complex initiatives. When outsourcing or purchasing off the shelf tools, get a proof-of-concept tailored to your use case before committing to long term costs.
    

Which new AI/ML-driven initiatives are you excited about that are underway at Lineate?

We are currently working on a product recommendation tool that discovers unique patterns in the intersection of historical purchase and demographic data. It will be flexible enough to deploy for a wide range of use cases, with exciting lift results already coming in for event and product companies. On the programmatic side, we are in the process of rolling out a bid optimization tool which greatly improves revenue for SSPs as well as publishers.

Learn More: How TechStyle Uses Machine Learning for Personalization: Q&A; With Danielle Boeglin

Could you share some trends or predictions in this space for 2020 and beyond?

We have finally turned the corner from ‘gather zettabytes of data on anything and everything and dump it into a data lake for some future analysis’ to ‘if your data isn’t actionable, don’t bother.’ This is the year that the theoretical possibilities will turn into bite-size changes that are truly impactful. Along with this mindset shift, companies are moving at a fast pace toward data-driven automation and machine learning in practice in strategic and target parts of the customer journey. I believe therefore that more brands will be getting personalization right at the most important parts of the funnel and will see the impact of this in conversions and revenue. Likewise, with many brands that lean into the building trust and stellar customer experience yielding strong and powerful results, many others will follow suit.

Neha: Thank you, Elizabeth, for sharing your insights on the impact of AI and machine learning in ecommerce. We hope to talk to you again.

About Elizabeth Gallagher:

Elizabeth Gallagher is the Chief Revenue Officer at Lineate, a NY-based custom software development company—where she oversees marketing, sales, and product development. Previously, Elizabeth was Co-founder and CEO of the award-winning ed tech company, Pixeldream, where she brought dozens of high revenue technology products to market for leading organizations including McGraw-Hill and Pearson.

About Lineate:

Lineate is the leading technology partner servicing companies in the U.S., Canada and other regions to help them go from myth to reality with custom software solutions. With 15 years in the industry, Lineate empowers brands of all sizes worldwide to build smarter products by harnessing and activating their own data. Lineate equips companies and business leaders with the skills and technology needed to manage the full customer lifecycle – from beginning to end – in one seamless platform.

Found this interview interesting, want to know more about the future of AI and machine learning in ecommerce? Follow us on Twitter, Facebook, and LinkedIn.