Leveraging AI for Marketing ROI: Dash Hudson & Automat

AI can be leveraged in a variety of ways, and it’s rapidly proving itself to be a crucial utility many marketing applications, offering benefits like better engaging consumers, leveraging data and improving both brand awareness and revenue. In this co-interview with Michaela Atkinson, Senior Marketing Manager at Dash Hudson, and Lucie Loubet, Demand Generation Marketing Manager at Automat, both providers discuss AI marketing solutions, the engagement they produce, how to generate more ROI, what the right channels are for growth and what unique features can maximize engagement.


Q: AI is useful to companies when it’s applied to a specific use case and solves a specific problem. Can you give a few examples of the best ways you’ve seen it applied to sales and marketing?

MA: In our own case at Dash Hudson, the challenge was pretty clear and a prime example of where AI can be effective. Marketing teams are always debating what content is best to post and when, and brand managers and other figures within a brand often chime in with their own opinions as well. While debate rages, time is lost and often the output of a team’s marketing efforts – in this case on Instagram – can be less successful than they should be.

Dash Hudson addresses this by allowing brands and media companies to leveraging a unique AI that analyzes posts and the engagement they produce based on data from the content of the image itself, to create a unified view of what performs best and to help marketers understand in real-time what is resonating with their audience.

With this in hand, brand marketers save time and see the most engagement possible from  Instagram, with a live beat on what’s trending, what’s not, and what content to turn around. Brands can also do things like leverage user-generated content to further feed their community and better engage users – with brands like OUAI Haircare doing a lot of this to drive even more relevant content back into their community at the right time for them.

LL: The power that AI has today presents new opportunities and challenges. Consumers now expect more personalization and connectedness than ever before – we want exactly the right thing for us at all times, no matter how we choose to interact. Brands are devoting a tremendous amount of time, energy and money to their marketing efforts to match consumer expectations, and yet very often much of it is wasted – either on minimal conversion rates, or consumers who never tell them how they feel about their messages until they simply unsubscribe from content. To avoid these problems, brands need to give consumers the tools to achieve the most possible personalization and relevance to them, and AI can do that by efficiently processing the massive amounts of information needed to achieve that at scale.

One example of how AI makes this possible is technology like ASOS’ visual search tool. Consumers all have their own personal taste, preference and style, and the more a product matches that, the more likely they are to both purchase it and be satisfied with it. ASOS’ visual search tool leverages computer vision AI to allow consumers to upload a photo of clothing or accessories they like and find similar pieces within ASOS’ eCommerce store. Because ASOS carries more than 85,000 products, consumers can tend to be overwhelmed by the choices they face, but they can now instead use AI to seek out what they already want within ASOS’ offering, making it much easier to find and buy what they like.

From Automat’s perspective, we leverage AI to bring personalization to consumers through conversations. We help brands make each of their touch points (ads, websites, social media channels and more) conversational, so that customers can get the guidance, recommendations and responses they need in real time. Brands have the ability to engage in Conversational Marketing, Commerce and Support at scale by allowing AI to take the place of human representatives and provide one-to-one interactions with consumers that generate more meaningful engagement, unique consumer insights and provide sales lift that is harder to achieve through other personalization methods.

Consumers are really responding to this application because, in fact, conversation comes so naturally to us that when we are actually presented with the one-to-one assistance of an AI, we usually take it just to see what is offered, and often that allows us as consumers to discover something we didn’t expect, something we like, or something we then want to buy.

Just one example of this is Vichy. They deployed Conversational AI to offer a diagnostic and generate a recommendation for their consumers. This works great in the context of a skincare brand, which has multiple products and requires effort and refinement to find the right product for the right consumer, but it also works just as well when a bank’s customer wants to apply for a mortgage, or speak with a financial advisor about their student debt. In that way, AI can be very versatile as to the domain it helps see results.


Takeaway: AI has opened up a variety of possibilities for brands that they would’ve never before thought possible, simply because they would have cost too much in the past. AI allows brands to automate interactions with brands and communities to a degree that they never before would have thought possible.


Q: Both Dash Hudson and Automat have a big focus on brand engagement. How do you measure it and how is it different from traditional engagement KPIs?

MA: With engagement, there’s classic measures and then there are more total measures which Dash Hudson uses to give brands full, modern visibility on the impact their marketing is having with consumers. The classic metric is calculated by summing likes and comments and then dividing by the number of followers to produce an engagement ratio. Dash Hudson measures this on both a per post and a 12-week rolling basis to observe the overall trends and performance of a brand’s photos and videos.

Then, content can be divided up – whether it’s lifestyle content, user-generated, or based on other “pillar” categories – and engagement can be measured and affirmed there to improve the brands aggregate performance. Stories are another category where this same rubric applies, except a little differently. Engagement there can be measured by how often users tap forward, tap through or tap back on a story, or simply exit stories. High rates of exits and tap throughs are bad for engagement, but conversely tap backs to rewatch a story, as well as watching stories to completion are an indicator that the viewer is really responding to the content presented.

Dash Hudson also recently released the ability for brands to segment their Stories into pillars in the same way they would in-feed content. This enables them to deeply understand what Stories creative hits home, what makes viewers drop off, and essentially, how to produce better Stories content in the future and prolong viewership and engagement.

LL: We encourage our customers to think of engagement as true interactions, instead of one-way KPIs like seconds of videos watched, likes, or other traditional measures. We specifically focus on measuring conversational engagement: how many users open a conversation as a percentage of overall web traffic or ad views, how many questions do they ask or reply to, how many minutes do they spend chatting with your assistant.

These are important because they allow brands to form a clear picture of who the customer is, what they’re looking for, and what they need to move on to the next step in the funnel. For each client, the dynamics of a conversation and the likelihood of certain factors to influence a positive outcome are always different. In the case of conversational marketing, the length and variety of engagement that comes with conversations means brands can take more incremental steps to incremental purchase decisions, and this permits brands to more easily prove out the value of their deployments and bring them to scale faster.


Takeaway: Engagement can be measured in a multitude of ways, and it’s one of the few cases where each of those measures gives new visibility on how consumers directly interact with your brand and what makes people more or less likely to complete a transaction.


Q: How do brands define Social Marketing ROI?

MA: Engagement is a way to measure social performance, but the two real core outcomes upon which the cost of social marketing effort can be justified are revenue and awareness. Revenue is pretty direct in terms of measuring how much revenue you can attribute to your social efforts, or a particular initiative within your overall social media marketing program.

Awareness can more broadly correlate with the overall performance and loyalty a brand has among customers. Reach and overall impressions form the base of this, but then finding a community that truly engages with the content based on legitimate interest is critical to establishing a vocal base of support that then expands brand awareness even further. Seeing engagement from core groups of users that overlap with a brand’s core group of consumers really offers a lot of metrics that pull back into a brand’s overall performance and revenue.

Just to give this a bit more color as to how to measure and improve on this sort of definition, Dash Hudson’s own platform features a particular tool – LikeShop – that makes Instagram shoppable by establishing a store that lives in the bio link of a brand’s Instagram profile, and ties  content published on Instagram back to it. In other respects, awareness and impressions are huge for publishers and media companies, because that drives inventory and ad sales. Brands can work to tie awareness and visibility back to sales in multiple ways, but for media companies in particular, visibility itself is the commodity that they end up selling to other advertisers, so in that case it can never be underrated.

LL: It seems like, since social media is where so many consumers already are, finding ways to capitalize on awareness and generate sales is critical to getting the best returns from social media. This can’t be done in an intrusive way however, but something that is more delicate and finessed with expertise – is that really the case?

MA: I’d say unobtrusiveness is really a top priority, but almost anything that respects that principle can improve ROI – whether in awareness or direct sales. A brand’s social performance really depends on how well its underlying offering and values resonate with a community, but the community guides much of that experience. In the same way, if you offer those audiences experiences that appeal to them – whether it’s in how they access e-commerce, how they discover a brand, or how they are empowered to share their story of a brand – almost all are likely to yield outstanding results.


Takeaway: Brands shouldn’t underrate the benefits of awareness generated from their social marketing efforts, but not without measuring it either.


Q: Why are consumers so responsive to Instagram? Can you give examples of brands that are doing particularly well there?

MA: Consumers love Instagram because it’s so visual and so easy to digest, and allows them to explore streams of content while treating each piece of content with its own art and attention, so to speak. In this process consumers can discover new products, personalities and communities, and engage with any person or brand that they feel like they identify with.

This, conversely, really puts brands in the driver’s seat, because consumers are onboard demanding content that will resonate with them. E! News, Youth To The People, Fresh, Everlane and REVOLVE are all great examples of brands and media companies finding their own voice in this marketplace.

E! let’s people get the latest up-to-the-minute celeb news – something they’re known for. Youth To The People gets to put out content that’s transparent about the ingredients and quality of what the brand offers, along with really creative content and interactive stories. Fresh keeps things sensorial and luxurious, Everlane makes it easy and attractive to shop, REVOLVE allows influencers to truly resonate and then leverages the experiences they produce.

LL: I know I’m guilty of passing through a lot of Instagram content without engaging, but every one of these brands seems like they’re really doing something unique to them.

MA: I’m always showing the love on Instagram! But yes, we all have fallen into that habit at times, and I think it’s a symptom of having so much “noise” out there, which is why brands need to ever more urgently find the best possible messaging, timing and community for their social media presence to stand out.


Takeaway: Knowing what resonates most with your community is critical, because when you build on what resonates you can tap into your brand’s highest possible growth potential.


Q: What can brands do to generate even more engagement?

LL: Ultimately, consumers set the terms on which they want to engage. Whether your content is about education, entertainment, guidance, advice, recommendation – whatever the goal may be – you have to find out what is relevant to people at an individual level and provide it on-demand. Conversational experiences are really the best way to do some of these things because they have the delicacy, tact and optionality that’s required to influence people subtly, rather than the risks that come with one-way messaging, no matter how subtle it is.

Within that, it means brands need to think about their voice and speak with personality and consistency to their consumers. These are the same kind of traits you look for from a brand on social channels, but in a conversational context, it can come down to the attitude that a brand’s Conversational AI or virtual advisor takes – is it friendly, is it cheeky, is it a tomboy, is it a bro – these are personalities that serve to identify the way the brand looks to serve its consumers, and this is critical. No matter the channel – display ads, Facebook, Messenger, websites, voice, anything – the brand needs to know its own voice and keep it consistent to get the best response from consumers.

MA: It makes me think – the way you make people feel, whether it’s talking directly in a conversation or speaking through a post and a caption, is what makes people care about a brand, and it can be very intangible. Brands that master that are the ones we care deeply about, and with AI brands have more opportunities than ever to make people feel a real connection, a connection that is just for them.

LL: Very true – brands can never underrate how they make people feel – this is a real source of truth as to whether they will actually be loyal customers, or simply fall off another brand

MA: Does it integrate with Instagram Direct?

LL: Soon enough! And actually, because we’re already on Messenger and Facebook recently announced a focus on synchronization between its messaging products, it seems like that integration may be more universal for consumers than previously expected!


Takeaway: Brands need to develop the right voice, no matter the channel they use to engage, and the best way to leverage that is with the right team of experts and technology platforms to guide them to their best outcomes.


Dash Hudson and Automat are both providers of AI-powered marketing solutions that allow brands to better reach their consumers, engage with them and drive revenue. Dash Hudson leverages AI to drive deep insight and understanding of Instagram audiences and content performance to enhance engagement and ROI on visual marketing channels. Automat leverages AI to build virtual advisors that provide guidance, recommendations and more to consumers, generating multiple minutes of engagement, unique consumer insight and substantial sales lift as a result. Contact Dash Hudson and Automat to learn more.

Share this post: