The third episode in our Reinventing B2B podcast series explores innovative ways marketers can get the most from their data, including:
- how to recognize good data when you see it
- ways to assess and evaluate the best platforms and solutions for you
- not to mention a sneak peek into what the future holds – think AI and cookies.
Listen in, or read the transcript below.
Podcast length: 36 minutes
Transcript reading time: 25 minutes
James O’Flaherty: Hello everybody, and welcome to episode three of Reinventing B2B. And today we are going to be getting into something that transcends everything we do – how we manage the massive amount of data we have to consume in B2B. So, welcome to episode three: The Anatomy of Analytics. It’s hosted by just Global, the global B2B agency of the year.My name’s James O’Flaherty, I am the EVP of Growth at Just Global.
We are talking about the challenges of data in B2B marketing and what it presents – the common needs that we’re seeing, how we think we need to address them, and the role of ad operations. And this is something that we really want to talk about today – there’s so much that happens that goes into having the right data sets to make the right decisions. We’re also looking at how AI is influencing everything, and what exciting things are around the corner. The other thing we want to explore for people like me, who think we’re marketers but definitely know that we’re not data analysts, is how do we know when we’re looking at good data, and how do we know when we’re looking at the right kind of data visualizations?
I am very excited to be joined by two of the people at Just Global who I admire the very most – our EVP of Global Analytics, Emilie Lee, and our Director of Ad Operations, Kat van Biene. Thank you very much for both being here. We always start with a fun question.
So, this is the fun question! Except for Reinventing B2B, this podcast, which is amazing, what would be your second favorite podcast in the world? Emily?
Emilie Lee: It’s a tie between the CMO Podcast and How I Built This with Guy Raz.
James O’Flaherty: Kat, how about you?
Kat Van Biene: I’m going to have to go with We Can Do Hard Things with Glennon Doyle, and that kind of goes over all the hard things we do on a day-to-day basis – loving and losing, caring for children and parents, friendships, etc. It’s a great podcast and a daily mantra I have – I can do hard things.
Even if you feel like it is the perfect dataset, or platform, or channel, always make sure that you are keeping your eyes open to other opportunities as well. Because even though that platform or channel or data set might have seemed really amazing in 2018 and 2019, it's probably not as amazing anymore.
James O’Flaherty: I love it. I want to listen to that one, and hopefully everybody else wants to listen to it, about 35 minutes after they’ve listened to this. We are going to go straight into it because we have got a lot of good stuff to cover off today.
So, we always start with a fact or data insight, or a report that we’ve recently seen, just to guide some of the key parts of the conversation, and the ones that we’ve got for the start of this conversation are around analytics. So, 42% of data analysts and 40% of marketers cite manually wrangling data as their biggest challenge.
What does that look like, and how do you think that it’s evolved over time, Emily?
Emilie Lee: Well, first of all, I’m not surprised. There are so many different data platforms and data solutions popping up left and right, and there so many different taxonomies that exist. It’s really difficult to figure out how to wrangle them all and how to use them.
In terms of aggregating the data and visualizing it, there are so many different solutions that have come about. There are some classic ones like Tableau and Datorama under Salesforce Marketing Cloud. For a more enterprise solution, there’s Power BI and Looker, which are quite popular amongst our client base. And then if you want more account-based marketing, there are solutions like Octane 11 as well, which have excellent solutions, and these are all important for different use cases.
What is really important in terms of wrangling data is ensuring you have a clean taxonomy and that you have that proper QA in place. Genuinely, having someone like Kat on the team, she is our superhero and is saving the day constantly when it comes to QA, spotting text issues, and troubleshooting. She’s our star, so I’m very grateful for that.
James O’Flaherty: Nice. Thank you. Kat – what about you? What would your thoughts be on this?
Kat Van Biene: I think that’s why we chose the whole analogy of the human anatomy, because everything’s connected within our bodies and also within an agency. I feel like I can go on and on and choose many different analogies, but when it comes to AdOps and analytics, and how we work so closely with them, and together, I think a common one is how a weak core can cause lower back pain. But a strong core is crucial for your overall health and functionality. And that’s kind of how I feel about being on the analytics team and working so closely. We are crucial to improving the health of the agency and boosting performance through our execution of visualizations. And for my team, I strive to provide support to the analytics team, to the other departments, to take away those day-to-day pains, and just help coordinate and take things off the analytics team’s plate.
What is really important in terms of wrangling data is ensuring you have a clean taxonomy and that you have that proper QA in place.
James O’Flaherty: I love anything that has an analogy – it makes it easier for somebody like me to actually get it. You talk about how you work, but I suppose the thing that I have on the other side is how I understand whether what I’m looking at is right, and whether it’s good. And I think I’ve had that in the recent past, trying to understand how we even go about evaluating an analytics platform. Emily, what’s the right approach here? What are the things that you should be looking for to understand what good looks like?
Emilie Lee: I think starting with: what is it that you are wanting to get out of it? What’s the objective of the evaluation? That will start to lead you down a path. And if you’re thinking about a specific data set that you’d like to use to justify perhaps a channel or a tactic, understanding why you want to use that tactic and channel in the first place. What is the KPI? What are the overall objectives?
Want to subscribe to Reinventing B2B?
If that dataset or that platform seems like an incredible opportunity, and it’s everything and more, it’s really important to beta test it and figure out what could go wrong, what are the limitations, what’s the cost of it? Were there any errors? Did it ultimately justify all that effort to put that dataset or to put that platform into place in order to get you the information that you needed and prove whether or not that objective was successful or not? If it didn’t and if it was a pain, then you really need to figure out a different data source or a different channel or platform.
I would also say, even if you feel like it is the perfect dataset, or platform, or channel, always make sure that you are keeping your eyes open to other opportunities as well. Because even though that platform or channel or data set might have seemed really amazing in 2018 and 2019, it’s probably not as amazing anymore.
There are likely other solutions that exist, so please always continue to evaluate what opportunities are out there because especially in the world we live in right now, it’s really an ever-evolving industry.
James O’Flaherty: How do you guys keep up? Because you talk about 2018 like it’s the ancient past. I’ve got grey hair now, and I had grey hair then. But how do you keep up with all the different trends, because I’m assuming that there are different ways that you would even evaluate something from five or six years ago.
Emilie Lee: Yes, the way that I personally keep up to speed on the different platforms and opportunities that are available is through podcasts. I might hear something interesting that is up and coming that I should look into. Also, our partners who we work with are coming forward and telling us, hey – we have this new better opportunity available.
In B2B marketing we say relationships are at the core of everything, and keeping a really close relationship with our partners allows us to take advantage and even know about those data opportunities so that we can make sure that we’re taking advantage of them first.
We’re always in evaluation mode. We want to make sure we’re doing the latest and greatest, and driving business growth for clients of course.
James O’Flaherty: How about you, Kat?
Kat Van Biene: Yes, I definitely second what Emily says. I really lean on our partners and our different vendors. I personally have Google Alerts set up to alert me of anything new, for example, like the cookieless feature. But a lot of times I look at those alerts and read the article which summarizes the same thing that we’re hearing. So, what I like to do is actually contact my partner, whether it be ads for TradeDesk, to get their insight, learn more about it and go from there and evaluate to see what’s the best way to move forward, to test.
James O’Flaherty: I want to move on to another statistic because, I mean, it’s analytics, so I need to throw numbers at you. Marketing analytics are responsible for influencing 53% of marketing decisions, and that’s according to a survey by Gartner. What do you think that says about the way marketing is working today and what marketers need to know about those decisions?
Emilie Lee: I think 53% makes a lot of sense. Marketing decisions are being driven by marketing analytics. We want data to be driving what we’re doing moving forward. What I also hear in that stat though, is that 47% of decisions are being made without data and analytics being a part of the decision, according to Gartner. I think part of the reason why that 47% exists is because of the issue with wrangling data. That said, we are living in the time of AI. It’s just exploding everywhere. All these companies are actually digging into what they can provide, and determining if they have the appropriate infrastructure and talent in place to take advantage of AI. And that way, they’re able to drive more decisions faster, operationally, efficiently. There’s so much opportunity.
What I think is going to happen is if we were to do this podcast a year from now, and we were to look at that Gartner stat (which hopefully will be released again next year) I think that 53% is going to be more like 65% to 70%, maybe more. All because of AI exploding, and all the opportunity that’s available from an analytics perspective – warehousing the data in one place and being able to prompt AI to make all kinds of decisions really quickly and allowing us all to be more operationally efficient.
Marketing decisions are being driven by marketing analytics. We want data to be driving what we're doing moving forward.
James O’Flaherty: I love that we’ve got predictive analytics of the data that we’re using to talk about analytics today. Kat – is there anything that you’d want to add to what Emily just said?
Kat Van Biene: She pretty much summed it up perfectly, but people are torn about AI and what it means for our industry. I’m personally very excited and I believe we really need to continue exploring the different developing technologies out there, and continue testing them. And I think it’s just a very exciting time for our industry. I know it can be scary for some, but as Emily mentioned, I’m curious to see in a year from now, what it will do for these stats that we’re seeing.
James O’Flaherty: We’ve got a whole section to talk about AI, which is where we can get into the real detail of that. We’ve been talking about how things work, and we obviously work in it every day – you particularly live and breathe it because of the way our clients work, and what they do – but I think one thing that I would love to get a clearer understanding of is what are the primary differences, and what are the common challenges when you think about B2B versus B2C? I know you both come from backgrounds where you’ve worked across both. What do you see is the biggest difference, and maybe the biggest challenge you have with that change?
Emilie Lee: I love a good B2B and B2C comparison, so I love that this is coming up. As you mentioned, Kat and I both came from a prior agency where B2C existed, and I think we were both in B2C-land a little bit more previously to that. Some of the common challenges in B2B versus B2C? My mind goes to the buying mindset, the buying cycle length, the marketing channels leveraged, and the tracking required. And all that goes into the analysis of the outputs, so that that we’re able to actually analyze.
Starting with the mindset, B2B versus B2C – B2C is much more emotionally driven. It’s going to be a faster buying cycle. B2B, as I mentioned a little bit ago, is about relationships in order to nurture that leap through, and make it happen.
Let’s bring in a couple of examples. B2C – let’s say I want to purchase Apple AirTags, I’m staring at one right now. So that’s a fast buying cycle, right? I am going to hop online. I’m going to search for the cheapest AirTag ,hopefully it’s not a terrible AirTag, but likely the cheapest AirTag, so I search, go to the website, and purchase.
Because that buying cycle requires hitting that audience or committee so many times, and the type of channels that we leverage, including content syndication, are a little bit more difficult to track, it's going to be far more difficult to prove or justify some of the upper funnel tactics or channels that are driving that conversion once it actually comes through.
That’s a really fast buying cycle, pretty emotionally driven because I want it quickly and it’s very easily trackable. I also saw an ad yesterday, a programmatic ad – the path to conversion there was seeing the ad, searching for the AirTag and then purchasing the AirTag. Your brand to revenue tracking process is pretty fast and easy to measure there.
With B2B we need to advertise to an entire buying committee of decision makers – those who would move forward with particular software or a tag. It’s going to take some time as we all know, likely six to nine months if not more. And because that buying cycle requires hitting that audience or committee so many times, and the type of channels that we leverage, including content syndication, are a little bit more difficult to track, it’s going to be far more difficult to prove or justify some of the upper funnel tactics or channels that are driving that conversion once it actually comes through. And it’s difficult to track that lead flow process at times, again, depending on the channel. So ultimately the solution there is really leaning in and understanding what each platform can provide, and then prescribe the appropriate measurement solution.
So, there are more complexities. B2C is easier from a tracking and analytics perspective, B2B is just more difficult because of the channels, the mindset, the process. You might lose some tracking in that six-to-twelve-month window. If I think about the easiest comparison with B2B and B2C, it’s likely auto. Normally people are not going to emotionally purchase a car, let’s hope not.
I don’t think many people I know would do that and it takes time. It takes me months to figure out what kind of car I want. And it’s difficult to track once you get to the actual car lot and purchase that car. So, if I were to come up with a similar sort of B2B versus B2C vertical, it would probably be auto, and just the complexity between the two.
James O’Flaherty: So, it sounds like B2C has massive datasets, but short buying cycles. In B2B we have massive datasets and long buying cycles, so we have even bigger continual data sets. I’m not going to be like, B2B is better than B2C, but does it make it more rewarding for you over the course of time to have to stitch all of that data together to come up with the right story?
Emilie Lee: Yes, the level of knowledge that we need to have on the analytics side in every platform and the different reporting capabilities that are possible is high. We’re genuinely having to prescribe the appropriate solution measurement framework based on the channels that we select, that we know are what’s best to deploy campaigns. Stitching all of that together requires a ton of knowledge, so when that data actually comes through based on that framework we recommended, there’s always an evolution process – we’ve been doing this for quite a while so I’d like to think that we get it pretty well right in the beginning. But we evolve. The data has to tell us what we should evolve to, but it is really rewarding when that lead comes through, and we’re able to nurture it or increase the pipeline velocity with deal acceleration, and optimizing there. It’s harder on the B2B side, but definitely more rewarding.
James O’Flaherty: So, what are the common things that make it hard to get it right?
The level of knowledge that we need to have on the analytics side in every platform and the different reporting capabilities that are possible is high.
Kat Van Biene: Definitely tracking from an AdOps perspective. I was scared to move away from B2C as it was pretty straightforward in terms of tracking, but I’ve found B2B to be rewarding as we come up with these complicated measurement frameworks and strategies.
But translating that into a tracking plan and executing it is not as simple – you know, tagging the landing page, the cart page and the thank you page. It’s about finding the user engagement and not just basing off CTR. Does the user spend X amount of time on a landing page? Does the user scroll down to the bottom? Is the user interacting with the video, filling in the form? There are so many opportunities.
So yes, it becomes a little bit more complex in terms of tracking and then also trafficking too, obviously the AdOps handles – there are a lot more placements because we’re really trying to get that granularity. But I love a challenge, I love the details and seeing it come to fruition with these beautiful visualizations that the analytics team comes up with. It’s just amazing.
James O’Flaherty: We get a lot of questions currently from our clients around Google and what a cookieless future means. Kat, can you talk to us about it from your perspective?
Kat Van Biene: This is something that Google announced back in 2020, so it has been creating a frenzy since 2020. I notice it comes in waves and I’ve found myself typing up new articles for the past four years. But as a quick summary, Google is trying to phase out third party cookies, which I think most of us know. And what some people might not know, is that first party cookies are staying – they are not going away. So, it’s not necessarily a cookieless future. I think that’s a common misconception.
Cookies are staying. It’s just third-party cookies and our capabilities to track users’ activity across different websites. But we can still see what they’re doing on our actual sites and with each announcement that Google makes – their latest one was their 1% tracking protection test they rolled out across their Chrome users.
I think that’s what created this next wave, this current wave of frenzy. Our clients are asking ‘are we’re doing enough? Are we prepared?’
I think that it can be scary, especially for our clients. I think it’s just human to fear the unknown, because truly no-one quite knows what’s going to happen.
And as of right now, it really relies on the UK’s Competition and Markets Authority, the CMA. They’re one of the main reasons why it’s been postponed for so long, because they’re evaluating the impact it’s going to have on revenue and the industry. They’re evaluating the impact on privacy and just evaluating the competition too. So with this 1% tracking protection test they have been really evaluating the initial results.
It’s already facing some challenges where they’re actually at a standstill, where Google aren’t allowed to move forward with the phase-out until they address some of these issues. Google’s made it very clear that the Privacy Sandbox isn’t going to replace third-party cookies, but it needs to be able to fill in some of those gaps that these third-party cookies going away will create.
That’s a current status now – everything’s at a bit of standstill for the next 60 to 120 days. And the CMA is asking everyone in the industry for feedback during the standstill. That way we can decide as an industry how to move forward. Again, I think it just comes back to the fear of the unknown.
Do people want to invest ahead of time in resources that can cost a lot of money to develop these technologies that can better prepare us? Or do we just kind of sit and wait and trust that these bigger players in the game are going to provide these turnkey solutions for us? It’s a crazy time right now, but I’m still not worried.
What we can be doing now is testing, and understanding what's happening when it comes to performance and whether or not these solutions are going to work.
Emilie Lee: One of the questions you asked earlier, James, was how do we know we’re using the right data and how are we making sure that we’re staying on top of the different data sources and different platforms that we use? And this is a really great example of where brands and their agencies need to be working together really closely to identify different tracking and measurement opportunities with the potential cookieless world coming through.
For example, you’ve got The TradeDesk’s Unified ID 2.0, LinkedIn has a conversions API, Facebook has a conversions API as well. What we can be doing now is testing, and understanding what’s happening when it comes to performance and whether or not these solutions are going to work. That fear of the unknown will start to dissipate a little bit once brands and agencies, when they’re executing their campaigns as they’re testing this out, know what will actually happen.
It’s small scale right now, but we have to start somewhere. We’ll give them a flavor of what could happen, and the different data sets we could work with once those cookies potentially go away.
James O’Flaherty: And talking of flavor, you have an analogy for this that I think is worth bringing up here.
Kat Van Biene: Yes, I always have a good analogy. It’s such a complex topic that I like to find these analogies to help better break it down and make people understand. For me, and I think for most people out there, when you’re preparing for your first child there is a fear of the unknown – even for your first pet.
But it’s about embarking on a new adventure. You can read the books, you can read the articles out there, you can do all of that, and then it happens and you’re like, ‘Whoa’. And I honestly think, just like in parenting, you rely on a village of people. You rely on your partner, and you rely on your friends that have kids.
That’s the same thing here. You need to rely on your village of partners and professionals to get you through this ‘scary’ time. And so, leaning on your partners like The TradeDesk, adserve, your Facebook and LinkedIn reps to see what we can be doing to prepare for this.
I wanted to include this quote from Julius, from Analytics Mania – I subscribe to his newsletter – as he really breaks things down and I always look back at this quote and I start to feel a little panicked about a cookieless future. And he says: “What should you do about the phase-out of third-party cookies? Not much. That is Google’s, Facebook’s and other vendors headache now. They will probably rely more and more on AI and data modelling and personal information of your visitors. Time will show what will happen and when more is known, trust me, there will be a buzz in the industry and you will be informed.”
That always reassures me that it’s going to be OK. We have some great partners, great vendors, and we’re in constant communication with them.
James O’Flaherty: This has been a topic of discussion for four or five years. Has there been an impact, because there’s been a lot of preparation for something that hasn’t happened a lot of times? Are there common things that have kind of gnarled up a little bit because of this in the last few years? Or is this just something that everyone is getting on with until it’s an actual thing?
Kat Van Biene: I do think that, as Emily has mentioned, each platform has their own solutions in place. Whether it’s the Convergence API set up with Google ads, there are enhanced conversions that utilize the first party data that we’re collecting. Right now, we’re still using third party data as much as possible, but we’re supplementing it with this first party data and all the other alternative solutions that they’re recommending we marketers really explore, such as AI technology and contextual targeting.
There are just many developing technologies these days and so many endless opportunities to explore. It’s actually pretty exciting.
You need to rely on your village of partners and professionals to get you through this ‘scary’ time.
James O’Flaherty: You’ve mentioned AI twice now, so it’s only fair we get to all of the stuff that I know you want to talk about from an AI perspective. So, we have the rise of AI. Can we talk about the ways the industry, and us, are looking to utilize AI within analytics?
Emilie Lee: With AI and analytics in particular, there’s a ton of opportunity. I think what first comes to mind and the lowest hanging fruit when it comes to AI is efficiency, understanding what can be done faster with leveraging AI.
Uploading a very basic data set, of course without any API, we have to make sure that everyone is compliant, asking it what sort of questions would a CMO or a marketer, depending on who the audience is who we would be speaking with, what sort of questions would they ask, and then asking the AI, prompting it with those questions to see what sort of output is provided is a really great way to spark curiosity, some inspiration, some innovation, and see what sort of output comes out.
Something that’s really easy to do is upload a brain dump of insights either from your specialists, your channel owners, your own analysis, and then asking it to rework some things or what additional questions, would a CMO have on top of these insights that we’re polled, how can we then answer those? What are some hypotheses that come from these initial outcomes that we’ve uncovered? Day to day a lot of our data visualization is actually powered by AI with harmonization and just overall data integration that’s happening via the different API – APIs that are coming into the solution that we leverage, which is powered by Salesforce Marketing Cloud.
So that happens day to day. As far as the industry too, there are some initial marketing analytics AI solutions that are coming about which are really, really exciting, where all of the data might be brought into an ETL service and then pushed to a data warehouse.
James O’Flaherty: Can I just ask, what is ETL?
Emilie Lee: ETL is a solution that extracts, transforms and loads your data. This is the wrangling of data that marketers have a difficult time with. You have to bring all of your data into one spot and harmonize it, make sure it’s all normalized and speaking to each other, in order to then be able to analyze it. And then what happens next is you can bring that data into a data warehouse – Big Query, Azure – and then AI now exists that can prompt the data in these data warehouses. And it’s not just limited to performance data anymore, it’s also audience data.
There are all sorts of different data sets that can be connected into these data warehouses that we can then prompt. And then as far as the industry goes, there are some initial marketing analytics solutions that are starting to pop up, which is really, really exciting because it’s not just about marketing analytics in terms of performance, it’s about leveraging marketing analytics performance as well as layering in audience insights, intent data, being able to do some media mix modeling simply by querying your AI, which then leads to the data warehouse, which is what’s aggregating all of the different data sources together.
So, we’re starting to get into this world where rather than having these multiple data sets through performance and a planning standpoint come together and take hours, if not a week to build, it is a query and given to you within seconds, which is incredible. So, tons of opportunity there. We’re in a really, really exciting time.
Day to day a lot of our data visualization is actually powered by AI with harmonization and just overall data integration that's happening via the different API
James O’Flaherty: This is blowing my mind a little bit. How much of it can be automated, how much of it needs human intervention or at least human oversight? How does that work?
Emilie Lee: We definitely still need the humans, absolutely. We still need the people prompting the AI. We need to have someone submitting that request for the information. I was listening to a podcast the other day that was mentioning how editing the initial AI output is still quite a challenge, because while you might be suggesting a very simple edit, that output is going to be given to you all over again, and when that happens, you might get a completely different output than you had before. And so that’s not very helpful.
The other thing is there’s inherent bias that’s happening with the output of these prompts. You have to make sure someone is there and ensuring that it’s compliant and that we are continuing to train the models appropriately, so that inherent bias doesn’t exist as prevalently as it does today.
James O’Flaherty: You said that we still need the humans, so I just want to thank you for that. We’re still required is the summary of that point.
Emilie Lee: Yes, we’re all still going to have jobs. Fret not.
Kat Van Biene: Also I think for analytics and AdOps, AI is something that’s embedded in our day to day already. I think we’ll find is as this year goes on, as weeks go on, it’s going to be embedded in all our day to days. And that’s just the reality. And it’s nothing to be scared about. I think it’s important to understand it, to test it out and to understand the benefits of how we can use it to make our lives easier, to become more efficient. It’s pretty funny that there’s an AI that provides an AI prompt – it generates what you will prompt your AI to do. And I use it all the time because sometimes that’s a hard thing to do. Yet again, that’s the human being behind it. You still have to prompt it in a proper way to give you a desired output.
And even if the initial output isn’t great, you continue testing it, you give it a different prompt to perfect it. But that being said, there are definitely still some glitches in the system. I also have noticed through many tests of different AI platforms that sometimes they don’t have access to the latest data, like Chat GPT, which I think is solely up to the end of 2023.
I can’t necessarily rely on it to know what’s going on in 2024, for example, with cookieless tracking, like what’s the latest and greatest? So, there’s still a need for the human touch behind this. And especially when it comes to QA, as that’s something I’m exploring, to see how I can incorporate AI and animation into the process.
But for right now, it’s still a very manual process from our side to make sure that we’re executing campaigns flawlessly.
James O’Flaherty: There was something you mentioned around prompts and AI, and creative. Our creative team has done a lot of work around the right prompts to use all the different tools that they’ve got available to them. I know that there’s innovation that’s actually moving that forward as well from a dynamic, creative perspective. Can you talk a little bit about what that actually means and where it’s going?
I think for analytics and AdOps, AI is something that's embedded in our day to day already.
Kat Van Biene: Really 2023 was the year marketers started experimenting with what’s called generative AI for ad creative. Essentially for those familiar with Dolly, it’s like the text image where you prompt the tool to generate an image based on your description. And it’s pretty fascinating. They just came out with a new one called Sora which is text to video, which is just absolutely fascinating, slightly scary, but in a in a good way. But I think 2024 is when marketers are really going to dive into these experiments. And actually, I did see a stat out there from Ad Exchanger where in the next 6 to 12 months, about 88% of marketing leaders will either have generative AI implemented, or have piloted, or have an active plan in place to test these use cases. So it’s pretty remarkable.
I’m curious to see personally how it’s going to impact dynamic creatives and DCO, something that I have been heavily involved with in the past where a lot of the process was very manual – pulling in the different content, images, and feeds. I had PTSD from the setup of it. So I’m personally very excited to see, and work closely with our creative team to see, how we can incorporate these AI tools.
There are a lot of glitches there. You can’t just rely on it to display these ads in real time. It’s not quite there yet. Also, the integration is costly. And so I think that’s where we’re going to see a lot of these different companies pop up to provide these kinds of turnkey solutions, and the integration between the creative platforms and these AI technologies. I’m just very excited to see what’s ahead, and excited for our creative team too, to see what they can explore and test, especially with the personalization and the impact third-party cookies has on it. It’s all tied together. It’s all the human anatomy, it’s all connected – without the third-party cookies, which they use a lot to power these dynamic creatives, because they’re based on a person’s location, the weather, the time. So how is AI going to fill in those gaps when those third-party cookies go away? And how are we going to be able to re-personalize these ads to our users and to drive success to our clients with what’s ahead? And it’s really all exciting.
Emilie Lee: One of the areas that I’m particularly interested in on the DCO side, or the performance creative side is the potential savings we could get from a cost perspective.
So years’ ago, whenever Kat and I worked together with DCO, it was pretty expensive. The CPM was more expensive, therefore the cost per conversion or cost per lead was more expensive as well. That said, with AI, we’re on a path of a less expensive CPM. We would have to go through this process of saying, well, if we were able to get the conversion rate up, you know, ten percentage points, that certainly justifies the higher cost of the CPM.
But it wasn’t an easy discussion every time because if you were to compare a conversion rate, even if it was higher to a programmatic display conversion rate and the CPM there, it’s just very different. So I’m very much looking forward to seeing what the CPM is going to look like in the cost associated with this.
James O’Flaherty: That’s a positive way to end all of that – all of the innovation and performance that’s going to come from it.
I want to say thank you very much, to both of you. I really appreciate you giving us the time and also going into the detail of that. There are a couple of things that I was going to signpost as well. In a few days’ time, we’re going to create some content that’s going to be available to download on the site in conjunction with one of our partners, Octane 11, which talks to some of this. And the other thing that I wanted to add to flag as well, you both contributed to a really good piece of content which was uploaded on our site around the cookieless future. We spoke about that today. So that link will be made available after this. But thank you very much, both of you, for everything you’ve put into this.
And we will be back next month with a new episode.
Emily Lee: Thank you so much.
Kat Van Biene: Thank you, James.
