”Perfect Attribution”

In the latest episode, we delve into the intricate world of B2B marketing attribution. This discussion, led by industry experts, navigates the complexities of accurately measuring marketing efforts, balancing short-term and long-term goals, and leveraging innovative measurement techniques. Their insights provide a comprehensive guide for enhancing marketing strategies and achieving more precise attribution in B2B contexts.

Guests

Alex Venus, Head of Digital and Web Marketing at Personio

Rupali Singhal, Marketing Analytics, EMEA and APAC at Twilio

Jonny Colclough, Senior Digital Demand Generation Manager at Scandit

Welcome to B2B Marketing futures. Today we have an exciting episode lined up as we explore the topic of perfect attribution. So how can we accurately measure the impact of our marketing efforts and ensure that every penny spent is accounted for? To help to unravel this complex subject, we have an amazing panelists today. So let's start with a quick round of introductions. Rupali, would you like to to start?

Hello everyone, I'm Rupali, I work at Twilio. I lead the international marketing analytics part at Twilio. I've been here for about three years and I focus on all kinds of marketing measurement challenges, from, campaign measurement to a more holistic measurement, which I lead through, media mix modeling at Twilio.

Brilliant. Thank you so much. Alex.

Hi, everyone. I'm Alex Vinas. I've, met Personio. Personio are a German founded HR software provider. Very exciting space of, HR tech. I've been doing digital marketing for over a decade now. Actually, interestingly or uninterestingly, I've been within a very specific part of B2B SaaS marketing within that HR function now for a few years So previously two personas with workday Pee con before that and a host of other companies, as well. My role in Personio as a head of digital marketing effectively. Within that we obviously have different tracking, measurement and attribution challenges. Some will be more like the website obviously, obviously on the performance, piece as well. And then also managing our POG sort of programs. So there's very much a broad selection of challenges. But also like really sort of fascinating areas around the space at the moment. So really excited to speak to everyone today.

Thank you so much. Jonny.

Hi I'm Jonny Colcleugh. I'm the senior digital demand generation manager at Scandit. I'm responsible for developing our multi-channel paid marketing strategy and also the inbound performance for the EMEA and APAC regions. I've been at Scandit for just over three years now, and before that, a selection of other companies in the B2B tech space. Throughout that time, measurement and attribution and it has posed a variety of challenges, which seem to be ever evolving in that time.

Thank you so much, Jonny and Tom.

My name's Tom Gatton, and I'm the chief executive of Adzact , and the co-host of the podcast with Joaquin.

Thank you. Tom. We've seen a shift towards making content really available. We know this is a trend. So the traditional method of using a white paper to get downloads, make a follow up call isn't working or isn't enough anymore. There's also the challenge of how to prove that the leads that we are generating, are generating revenue. What do you think are the core challenges with attribution in this context?

One of the main reasons why in the past you would use gated content so much was because it, added an element of measurability to campaigns. It was very clearly measurable, which I think when trying to communicate out performance to the wider business was quite helpful in trying to show that, success of a campaign or a particular piece of content. The shift away from it. We've now moving to, ungated content with friction free journeys. The challenge is it's much harder, actually, to measure. I think you have to be a bit more kind of nuanced with how you actually measure the success of not just the content or the entire program. Basically, I think that's the big challenge is trying to look at the big picture with the entire program now, not just one particular tactic or campaign.

I think the, B2B marketing has become more multi-channel now than it used to be once upon a time. I was reading a stat somewhere a few days ago as well, the B2B, ad economy has tripled in the past ten years. Like how much money companies are spending on marketing. That just means, we are bombarding like our customers have a lot of touchpoints. They've got digital touchpoints. They're hearing about us from email. There's not a singular journey for a lead, to become a customer anymore. So how do you make sure that you are giving credit where it's due at each point in the journey?

I think I definitely agree that we are at an interesting point where in general, I would say the measurement available isn't aligned to the tactics as much as they once were. I definitely agree as well, that there's a long tail of diminishing returns and everyone's doing the same things now. I think the one thing I'd add to what's already been said is I think there's obviously the operational shift that needs to happen and the content marketing shift. But I think the other thing that a lot of digital teams are addressing as well is the sort of the mindset shift and the organisational buy in that has to sort of take place as well. When we're saying, hey, 50% of the tactics that we're going to do probably won't be able to measure it that well. And that creates a whole process of having to internally market your new external marketing strategy. I think we're definitely at the pivot point I suppose. In general, all of these things are positive, right? I believe that, yes, maybe the E book into a lead nurture funnel was more effective five years ago, but it's also wasn't ever, that we had 100% ROAS on these sort of activities. Right? So I think we are seeing diminishing returns. But ultimately I think marketing now across B2B, but in general I think is going to be improving as a consequence of this. The end user experience is going to improve as a consequence of this. And I also think the profitability of marketing's programs will eventually catch up and improve as well as a consequence of this. So we're at a pivot point. But I think also like this is directionally a good thing for for marketing and the end user as well.

I think we need to learn how to think about B2C modes of attribution because, I don't think they're necessarily applicable. And I think that people have imported, B2C way of thinking, last touch attribution, for example, which you still hear lots of people talking or bemoaning the fact that various different channels don't allow this, and kind of trying to adapt the B2C modes of multi-touch attribution to B2B, there are many problems with this, but I think the main problem is that it's just designed for B2C, basically. But like one particular example is if you've got a salesperson contacting any person within this audience, you can't balance the two. People don't require the same level of evidence from a salesperson. They can say, I spoke to this person. And then three weeks later they emailed me and then they bought three weeks after that, you don't try and atomize it and say, well how can you prove the neuron, like moved from there to there in this person? And that moved their hand to click on the purchase button, the level of evidence required rightly is much, much lower than you'd have in the B2C, where if you want to track someone's progress through a website all the way to a cart and so they're just incompatible modes of attribution.

I've seen this where, people are desperate to try to keep things simple and understand why that makes sense. But then we'll look to implement a lead model that would make sense maybe for like a startup, but wouldn't make sense for an enterprise with a deal cycle that's, 12 plus months and where there's a matrix organization. And I think as well that wanting to have this. Okay. But how many website visitors did we have. And then what's the conversion rate to that. But that you're ignoring the intent, the segmentation, the filters that need to be applied onto this you may have a long deal cycle. You may need to buy in across a committee of buyers across your organization. And I think the other thing that I see as a challenge for this is a lot of this is focused on a lead or a contact level. Whereas I think there needs to be more of a shift to looking at an account level one because of the data issues that we have now. So look back windows, cookies, all this sort of stuff. But you need to be looking at how you're warming that account up over time. And I still think most measurement strategies are lead focused. And so they're looking at individual contacts. B2B has this great advantage that we can use account based marketing, can use to account based intent aggregation and we can know exactly which accounts we've served ads to. On LinkedIn, for example, you can upload a list of specific accounts. You can do it on adzact too. There are loads of account based marketing channels now, and you can get impressions and clicks by account. So you know who's going in the top of the funnel. And then of course you know who's converting. B2C doesn't have anything like that advantage.

Yeah but in B2C the buyer who comes in is the actual decision maker. Whereas in B2B the person who comes in is often not the person who you're seeing in Salesforce talking to the salesperson, and that's why, to Alex's point, we need to focus on account level rather than lead level because, the leads don't matter as much. That said, we don't put like in this world of focusing on accounts like a Twilio. We definitely focus more on account level now, but, we do not then focus enough on getting enough, elite enrichment or account enrichment with more contacts. That's the pitfall of focusing on an account level where we're not even measuring how many leads have we got engaged. So there's a little bit of a balance between the two that yes, you need to have multiple touch points within an account, but also you need to constantly think about how to enrich that account database as well.

It's a really tricky balance. It's like you're looking for these like really precise signals in a campaign or a program or an account, really high intent leads coming through or you just like, that's it. It's a really strong signal of buying intent versus softer aggregate signals at an account where it's trying to know, when can you use, some of these more kind of broader brush strokes and measures over time to say, look with this campaign and program like this account we're seeing in LinkedIn is showing engagement. And then asking questions of what is that telling us? So it's not 100% reliable. Oh yeah, we're seeing this account just engage. So let's do loads of outreach. But it's almost like just trying to like take it for what it's worth at times. How do you even define good account engagement. Right. What defines a quality account Engagement is seeing an impression of your, ad enough. Is downloading a white paper enough? In a company the size of Lego, for example, is someone sitting in, singapore downloading my white paper and me generating business in the US? Can I actually imply a correlation at the moment? You kind of do and in the multi-touch attribution models that we have in place, I've hardly ever seen people limited by regions, but especially in enterprise accounts, it's very hard to judge what a quality account engagement is. And do you account engagement the same.

It's so nuanced, I think, because sometimes you want these measurement models. They can be very helpful in guiding and be a giving you an engagement score or something like that. But then sometimes it just pays just to talk to the BDR or field team and just ask like, what's your take on that? What's the history of this account is this something which should be, dug into more? Is there any kind of follow up we can do with this account based on this? And it's, it's this mix of like quantitative and qualitative signals.

Have you both implemented self-reported attribution and what's been your experience with that alongside sort of the traditional types as in, how did you find us on that sort of qualitative, I suppose, and then sort of correlating that with self-declared form intent, that sort of stuff.

We've not added that in on our forms. But when we look at our multi-touch reporting, I guess what's coming up now is this showing in the last touch? But then a comparison it's always this question of what was the first touch as well now? Which I guess looking back over time, that wasn't the question. It was just like looking at the last touch, which I think is positive. But then sometimes in campaigns, we're almost looking beyond the first touch. How do we get into that first touch? Our campaigns are probably going to influence that But I think self-reported is massive help it by the looks of what I've seen people talk about and share.

I don't think we asked that question on our form anymore. And that's two reasons. One is we want to streamline we want the customer to fill three fields and get on with it. It's something that we never thought was too necessary. Second is, a company the size of Twilio has built enough brand equity. So it's very hard for someone to say, you first hear about us, it's a popular brand at this point, at least in the communications API business where people come to us from. And that's why we haven't. But have you found anything on it? From personio's perspective? We run it more as a pilot for several months to look at the correlation. So we default to first touch in Salesforce. And then we wanted to do a sort of a comparison there. We actually set up an AI script as well to basically normalize that data. Because obviously when you have a free text field, you get a lot of, random responses in there and no one has time to sort of piece through that. We did see that the match rate, like we obviously saw some of our sort of Dark Funnel activities. So we run a lot of we run regional podcasts now in each of the markets. All of the offline stuff did get representation. I actually think we saw a higher match rate between first touch attribution and self-reported than what I thought we were going to see. Which I dunno if that's a good thing or a bad thing, right. Because on one side it's like, well, oh, that brand spend actually wasn't going towards something or that that podcast didn't have the impact. But we did see a lot of that stuff getting called out. Do I want to run it as like an ongoing piece? I see it more like a brand lift study where or where you will run it on like a every half year to as part of maybe sort of a measurement exercise, or you'd run it for a quarter and then just as a litmus test to maybe after a big activation or a rebrand like we'll be doing at the end of the year. In general, I don't think it's something where it's going to be a fixture of MBRs or something like that. But in general, all attribution for me needs to have that post analysis as well. I sometimes think that, your earlier point that, we're looking at maybe linear attribution, but across those touch points, they're completely unequal, like one touch point required someone to, go halfway across the country. And then the other one is a website visit and it's like, well, we can't be looking at these in equal ways. And so I think all attribution, whether, there's a, first touch digital approach or whether you ask people how they heard about us or you speak to a salesperson, they're all just data points. Like none of them need to overrule one another. None of them, I think, is perfect. They all have their flaws.

I always have queried this with looking at these attribution models where you line them all up and then it would just be this and then we just peanut butter spread it across. How can we do this if we're lining up it all up. We've got all the data here. And then we just layer it across like that. I just thought from that point I'm like, this falls down a little bit for me there. It's one of the flaws. It's always useful, you've got to interrogate it for what it's really query it for what it's worth.

I think one thing that Alex, you said earlier about, it's good to check and see what your attribution model is telling you now and then with these experiments. It has been an observation, working in B2B for the past few years, that we are a little bit simplistic in terms of our multi-touch attribution models. So people are implementing either U-shaped or J shaped W shaped. Sure. I think all valid, approaches, but there are techniques to be more mindful about the weights that we give to each of those touch points. The Incrementality tests could be a way of verifying the weights that you give to each of those touch points. It's not just about like you giving 40% to first touch, 40% to last touch, and then 20% distributed everything in the middle, which is the classic U-shaped model. But maybe we give a higher, weightage to, someone going into an event where they have a proper conversation with one of the salespeople, and they are able to sell the product in person to them. But a lower weightage to an ad that they saw in passing on Facebook.

How do you decide those weights from a qualitative analysis or from a qualitative analysis.

So there's a lot of, machine learning models in place these days where you can, give it customer journeys that actually lead to conversion. And they help you calculate the weights, but those should be verified by incrementality experiments. So like, um, so you stopped doing something. Yeah, you could do some geo tests. So, for example in one region, switch off your Facebook ads completely and see if you actually see a drop in the number of leads that your attribution model is telling you predict that you are getting. Exactly. Or you could do, like that's a geo test, or you could do an incrementality test as well, where for a particular period of time you stopped doing something and see the kind of impact it has. Then you can sense check the weights that you're giving your attribution model or even use it for your, ROI calculation, that's not something we generally do that much, because all I think part of it is also because in the B2C world, they rely very, very heavily on marketing, whereas for each customer touchpoint people coming in. The B2B companies grow marketing reliance it becomes less because you have build all that brand equity and you have a big enough sales team to sell it. This is the direction for me. It's why see a lot of, companies now sort of building a centralized BI , function, building proprietary tools, maybe initially they would have their kind of core measurement within their CRM or something, and now sort of building something within a proprietary data model, because I think what I'm seeing is that the attribution models aren't sort of maybe fit for purpose for the measurement strategy that they're trying to do. But then also, I see this is slightly different around the scoring model and like what a best fit account looks like. And I think being able to tie those things together, like weight actions based on relative impact in terms of like how many customers it's bringing the right sort of fit of customers. I think all this stuff together, it requires a level of sophistication with the tech stack that I certainly think companies didn't have sort of five years ago. We're having like this kind of like BI catch up, which I see at the moment is, we need to change the metrics that we're looking at, the attribution models that we're using, but then also connecting this into like new user journeys as well, and then weighting these user journeys not on clicks and impressions, but on actual ROI and revenue.

I'm pretty sure skills as well. I meet people every week that have just come in to B2B from other disciplines because as you say, Rupali, there's so much spend now on B2B digital compared to 2 or 3 years ago. There's much bigger teams and many more businesses that are spending a lot more money. And and these people are going to have to learn this stuff. And frankly, the people that are already in B2B are going to have to reskill and rethink this. The sorts of skills that people might have in B2B in three years time might be quite different from the sorts of teams that we have today.

I'd agree definitely. A lot of, tools available now for lead grading, account grading, predicting next best product. All of these, have been getting better and better in the past couple of years, and definitely seeing a lot more data science talent coming into, the B2B space, building all of these, cool tools.

How do you balance short term goals with long term brand building efforts, and what kind of challenges you see in attribution related to that?

One of the challenges when looking from the attribution point of view of the brand building side of it. Under a lot of models your brand building activities won't necessarily get a look in. It won't appear in the attribute. When trying to actually measure these touch points, how does it factor into the attribution model and the ROI, if it's not being measured in the first place, then we're not going to it could actually be a valuable touch point, but the model will measure what the model is designed to measure or program to measure. It's tricky trying to sell the value of these brand building and not just showing the brand, but also showing I think the key is it's showing the value of your product and the brand so that people aren't just brand aware but when they actually are in market, they do think that you are the solution, which I think is the thing which we often get wrong where we can build brand awareness. Widespread ad campaigns and stuff. But then if it's not telling people like why you're better than the competitor solution, then the where attribution model you've got, you're just not going to get in there anyway. You should know what's working at the bottom of the funnel. I still think particularly in that demand capture level like analytics and reporting are strong enough now for for most decent teams that you should you should understand what's working and have a testing program around that. Outside of that, your focus needs to be on. Having an impact in that 99%, 98%. The other issue I see with and this comes back to kind of that, sort of oversimplification that I see sometimes I still feel like people look at marketing channels in isolation way too much now rather than synergies. On my side, I'm a performance marketer or run a performance team, and we're always advocating on focusing on brand versus performance marketing as our like our main growth lever moving forward, particularly with our maturity in the market, which I know sometimes is sacrilege. I should be asking for more budget all the time. But you can't have a high converting funnel with stuff with that awareness, not sort of growing it at the top. Right. So even as a performance marketing team, and we obviously do think that we have more of an impact, we know we have more of an impact. Further up the funnel, we'll look at metrics like, share of branded search. We can always see that correlation between growth of brand and growth in terms of performance down the funnel as well. But it is really clear to see that, if you just try and optimize that level and particularly given it all, it matters where you are in the market. But all of your competitors in general are doing the same tactics down the funnel. I think what I see most people really struggle to do is get that organizational buy in to change that, investment spend and, really sort of shift investments further up the funnel. But that's something that we advocate pretty much every single QBR. Right? Is it's the route forward. There's tactics and efficiencies that we can do across the whole funnel. But the main thing for me is, how can we sort of create stronger differentiation, how can we improve the user experience and UI and across our digital touchpoints? That's coming from the guy that runs the performance team. So I think it's clear that there's everything's related.

So refreshing to hear that because usually you always hear performance marketers talk about Google paid search. It's really cool that you, focus on the brand building efforts as well. And I also think like brand building is different for a lot of B2B businesses than it is for B2C, right? Like for B2C, it could be outdoor advertising, advertising in newspapers, TV, sure. But I think for B2B it's the Alex, you mentioned persona does community podcasts That is brand building. And you actually saw that coming through when you, put the, question in your form, how did you hear about us? Right. So clearly, that kind of thing works even for Twilio. Twilio has a huge developer evangelism team. All they do is go out in events and hold events to for developers, and we take them for free. We invite developers in, which teach them about the product, we teach them about a new capability.

It's not supposed to generate ROI. I don't expect to see those particular signups or actual revenue coming from those events. But this is how Twilio has sort of built its brand. It's a developer first company. Whenever I go to tech events, people see my badge and go, like, even if it's not a communications API kind of event at all, people see my badge and go, oh, Twilio. You know, as a software engineer, I used your product or, you know, people have experimented with it. And I think that's how Twilio has built a lot of its brand. So I do think it has helped the business grow significantly, even if we cannot directly prove its ROI through attribution.

There's two problems. Why? I still think companies are struggling to make the shift. One is it's really hard to do good brand marketing if you're not an established household brand with a clear brand identity. And I still think there's a misunderstanding around brand and marketing is there's a difference in my mind between sort of reaching people and obviously having that engagement with someone. So brand awareness versus brand affinity and doing brand marketing well takes the same level of testing and experimentation as what a performance team would be doing down the funnel, right? It takes a long time to understand that and perfect that. So I think that, most marketing there's been a legacy, I think, in B2B marketing of people just, shipping a JPEG with some stock photography and, ticking a box in terms of brand marketing. But to effectively do it well, it takes, real higher investment in design creative direction than most marketing teams are willing to make right now, particularly with the stuff that's going on in the market where we're at that pivotal shift. People are demanding to see stronger ROAS or stronger ROI to then say, actually, I need to invest more in, design visionaries and creative visionaries. It's really hard. It's really hard to do. And the other thing I see with brand marketing is people trying to overengineer the measurement strategy around it. I think sometimes you just need to go not just on gut and instinct. But yeah, we've share branded search. Yes, you can then blend that with some survey data once you've got an established baseline in terms of blending your quantitative data with some qualitative data, you can look into having that more of a media mix modeling strategy on like a quarterly or half year basis But some people are trying to look at brand brand marketing in terms of ROI and catching up on or every other week or a monthly. It just doesn't work like that. And I think people are still catching up. It also creates this fatigue when a CEO or CFO is listening to these presentations because it is unfortunately still a lot of vanity metrics with a lot of stuff where it has a very indirect correlation to the metrics that they care about. And so I think as a marketing team, you need to sort of change that narrative, keep things simple, show the correlation that you can. Basically just try and play catch up by increasing your investment into that brand program. I'm still more confident than ever that it's the right thing to do.

It looks like you, give a lot of importance to creativity and creative marketing in your company. From a performance perspective, do you collaborate actively with with the creative team? How you structure for example, creativity within your company?

This is one of the biggest changes I've seen in terms of the profile of our team, over the last sort of, five years or so when I've been in performance marketing, is it used to be a super technical field. And I'm not saying it's not a technical if anyone's listening like we do. Still we do still know data. But it's also being able to have people who are confident in creative testing. Where we used to have people who were running calculations, where it was manual CPC on search, whereas now I think the real growth lever, for me, the biggest growth lever in performance marketing is creative. Everyone runs search ads. You can, run scripts now, even with AI to automate a lot of this stuff. So the competitive edge for me is, reaching people with a video campaign or reaching people with, a high level of creative testing and being bold and that stuff takes a whole different profile and skill set. So which is what we're hiring for. So we now have designers within our team, but then also a super close relation to, an internal sort of scale team and also an internal sort of creative direction team. And then the other thing that we've tried to do is shift our investments a lot more. I still think, most performance marketing teams will be spending 99.9% of their spend on media versus enough on content, production design and real creative testing around this area. And we've tried to shift that a lot more. So to try to have a bit of a healthier ratio, obviously it is still mainly weighted towards distribution versus, that sort of design arm. But now when I look at performance marketers I guess I look a lot more to them as a marketer than maybe as a deep technical practitioner, particularly on the search side. And what I would have done previously, maybe that's just me, but I need people now who can obviously work cross-functionally really, really well, and sell that vision I'm trying to do. But also be confident writing their own copy. Be confident coming up with creative ideas. Be confident working in a room with creative directors and designers. And I think that's a big shift that I've seen over the last few years.

We've seen some performance teams with creative teams embedded in them, which we probably wouldn't have, but definitely wouldn't have seen three years ago.

From my background starting out, it was real focus on hard performance data and analysis of in the channel. If you're not building the ads and creative yourself, it's being able to look at like the composition and anatomy of creative and content to be able to just kind of steer and say, based on the data that we have got, we think it's this element which is driving the performance. This is a direction which we would like to try. It's being able to contextualize the why it's got engagement and why it's actually, performed better or worse than what you'd thought, which I think in the past it wasn't so much of a requirement because it was just more down to the, manual campaign setups and, run lead gen stuff like that, where it was now it is such a focus on creative, your creative just being on point and it's these intangible factors like salience, isn't it? And just being able to just stand out. How do you tell people that you need that, different skill set now?

New measurement strategy as well people analyze engagement rate in this aggregate way at a channel. What's our engagement rate. There's a lot of powerful breakdowns now within platforms. But you can, create your own custom metrics as well where you look at the hook rate. So how do the first five second hold someone? Where does someone drop off in a video on a YouTube ad? I still don't think a lot of people are having these sort of that level of granularity when they're looking at engagement metrics, because obviously engagement metrics isn't pipeline isn't revenue. People still would see it as a vanity metric or something quite fluffy. But if you want to test at scale, you need to understand why that creatives work not working just as you do with your trial conversion rate or your demo conversion rate, what's happening there. If you want to test at scale and invest in paid social creative, for example, and you're not doing that level of breakdown analysis on your creative, then you're probably doing it wrong.

And It takes a significant shift in the company mindset to advocate, to think about those metrics rather than because if marketing is not producing pipeline, the ultimate question is, of course your hook rate is fine. Your engagement rate is fine. Where is my pipe? That's why the measurement should be capable to prove the value of these things as well. The incremental impact of better advertising, not just advertising.

What are the best ways you've seen of doing attribution by quality of content or kind of elements of content? I suppose you would That's what you're talking about, right? These hook hook rates. I think, Alex and John might be able to speak better to the exact metrics, but I can just give an example, that we have seen internally of, brand advertising working that did not come through in last touch attribution at all. Last year, because, all companies had significant budget cuts, right? All the tech companies had, significant budget cuts, layoffs, etc.. We cut our marketing budget significantly in H1 of last year. We reduced our PPC budget by 70% in North America. But the actual impact on our leads was less than 5%. That told us two things. One is, that we had significant cannibalization we were paying for leads that you would have got anyway coming into the funnel anyway. But the second part of it, the reason the story is not as simple is because some of that budget was actually reinvested into LinkedIn advertising, but LinkedIn advertising, which was not optimizing for performance, it was optimizing just for impressions and, eyeballs. And what we saw is that and because of like that, advertising was focused on C X audiences like contact center audiences, which is not someone we were traditionally targeting. That wasn't our audience for our messaging API, for example. It's more for our contact center product, which is called Flex. But that actually brought a lot of leads into our, the media mix model actually showed that that brought a lot of leads into our funnel, not just for the flex business, but also for our messaging API business, our voice API business, everything So that kind of is a a good example where brand kind of advertising to audiences we don't traditionally reach, create a demand for us across all our products rather than just capturing demand, which is what I accuse Google search of doing

quite a lot and giving you a false sense of oh great, my ROI seems to be good, but really you're just cannibalizing on top of what you probably would have already got.

I completely agree with that. One of the things I find with coming from like a paid search background and like paid is when talking about brand is the better the brand advertising is, the easier paid search is. And it's like actually the less money you need to invest in it. So it's when you take that like holistic view of all of the channel activity, it's like that then frees up so much budget resource to put elsewhere into building demand. But it's like it's shifting that thing. This is one of the I guess with the attribution models, like the model will tell you like paid search is driving this and it's like it's not all the time. Sometimes it can be. And in some markets with really intense competitive bidding and stuff, you almost have to pay to be there, for your own brand because of the extent of competitive bidding, especially in tech space now. It is having that nuanced view and then just looking at well, if brand is firing, then like paid search doesn't need to work as hard for me, you can put that budget elsewhere and then keep feeding the machine at a different point. But that note won't necessarily show up in attribution modeling. And one of the ways we find that to really show that this is working is just looking at this linked in account data of like who we're targeting certain campaigns, who we are getting this engagement from. And then if it is just manual match back in, I'm just like the leads coming in and being are we just reaching these accounts in LinkedIn? A cross-Channel it doesn't have to just be paid. It doesn't have to just be paid search. It could be paid organic. Whatever's coming through the Serp and then just just match back and and just being like, okay, are we actually hitting the right accounts broadly. And that can just be a, like a pretty rough and ready way of doing it. I think it does actually have value in the middle.

It kind of goes back to that point I mentioned earlier on where people are obsessed with going, what's working for us? Is it paid? What is it? What is paid and not seeing these channel synergies to speak a bit more tactical on this point, we've now added a new custom metric, into our reporting to look at the synergies between paid social and Google. So the meta pixel actually collects all website traffic data, not just obviously the meta traffic data. So that will then be picking up traffic from Google. So we have a metric a go to Google rate to look at basically how much Facebook is influencing that paid search pipeline that's coming in. I've seen this as well when I've done some sort of freelance work as well, when I've been looking, at previously and working in more like a sort of a Shopify environment. And I've always seen that some of these social channels maybe are underweighted, particularly in terms of lower funnel conversion impact. But then if you're able to then sort of have that in there and say, 30%, 40% of of these people that went to Google, then actually we ended up from Facebook, going to Google, it really allows you to actually step back and maybe make less erratic and reactive decisions and strategy decisions around this. And I'm not saying everyone chuck your budget into Facebook, but again, it's always just making sure that you're making these decisions with all of the data there and and having these kind of custom views versus just relying on one attribution view, when you're looking at it. So yeah, I love that sort of metric that we look at internally and that channel synergy.

Very cool. So this is a metric meta gives you. Not out of the box. No. So it's by deploying the the meta pixel. It's then collects all of the traffic data. And then we use the referral source to see where those people are going. So to actually implement that in the platform you create a custom metric in meta. And so then you'll basically have like a go to Google percentage point and you'll see okay, well 30% of people went here, 20% of people went there. If you actually sort of step back and think about it like I have this even working the field that I am, if I see a good Instagram ad or a YouTube ad or something like that, in general, I'm not clicking the link one because, you know, everyone in the world is already following me with the retargeting ads, but also because just my natural instinct is, I'll go into the browser and I'll just search for it directly. That's in general my user behavior as a shopper. I know it correlates with my own behavior which is another signal. But yeah, we see that ourselves, even with B2B buyers, that they're doing that. I don't know if they're doing that because it feels more intuitive, or they're doing that because they're also scared of tracking codes. We can see this within the data. We bake that into our regular reporting now.

Do you think it works? Because now people try to restrict cookies altogether? I use adblocker everybody uses like most people use Adblocker. Do you think it still works?

Partially I mean, we know that, even by changing certain cookie consents on our site, we see like 30, 40% data loss. There's no full picture view across any digital advertising anymore. I would never take that, say 30% or 5%, whatever it is, go to Google rate as a picture of all spend that you're doing on one channel. Again, it's a signal to say that, yeah, we need to just make sure that we're actually not going to see that cannibalization. Like I said, I've maybe done this more aggressively when working in for startups. It's like, yeah, actually we're going to switch up all of our Facebook ad spend because we don't see a strong ROAS. And then straight away you then see, oh, actually, my CPA on Google searches has gone up drastically. And then, my performance on paid brand is suddenly just falling off the roof. And it's like, what what do we do? A month ago, oh, we switched off all of our paid social spend. And I think, yes, that's maybe more magnified in very, very short transactional funnel. But it still applies in this kind of like longer more inbound content marketing B2B funnel as well.

We'd love to come back to what we were discussing before about the importance of brand building and creativity. What are the kind of metrics and tests that you apply to have better brand, better creativity? If you could tell what are your secret sauce. I mean, some of the stuff that we'll do we have some quite kind of standard ad templates that we've got in the locker, which we can just use to. And then we test that against new creative sets where you'd say, look, this is something a bit more kind of out there, something a bit more trying to stand out. And then that when you measure these kind of like stocks, template sets versus the new creative sets, that gives you a good data point to show that, this is the difference and it might not always work. Sometimes you think, yeah, the stocks that's actually done better, which can be you have to again, it's that looking forensically of why do we think this happened. I think it's just having like a very thorough ad testing regime really just like quite structured. That's probably where I'd start, really.

I think the thing that holds people back is that scale, and they're really the only scale that you can really have, particularly in a B2B space, I think is in platform through performance ads. So to try and get that scale, I think what we wanted to make sure is that if we are going to sort of scale these creatives, we need to a way to produce them quicker. So we create them with like a, like internally, we call like a performance toolkit where we'll have these kind of consistent structures. Maybe it's product UI or social proof or call to actions, basically design tiles that we know work really, really well. And these can be video, these can be, stock video, whatever it is. And then what we then do is, okay, what are we going to test on this creative. Are we going to test the hook here. Are we going to test the new, sort of social proof point Is it a new product UI that we're going to test that's obviously aligned to a different objective and by only maybe changing like 1 or 2 sort of structural blocks of this ad, it allows us to produce considerably more creative at a scale which actually allows you to test, and so then we go, oh, actually, we want to see which value proposition point is going to work. Well, we have some headlines that we were thinking of testing. What we can then just do is then load them up at, at the front end of these ad creatives, keeping the rest of the creative consistent and then actually saying, what, the hook rate improved on this ad that's came talking about this product proof point or, we saw that the call to action was stronger by applying these logos, whatever it is, potentially terrible examples. But I think the thing that really holds people back is one being bold and brave when it comes to these creative things, that there's no point running creative tests if there's a lack of diversity in your creative formats, or if you're changing blue to green or blue to green, or I've seen, different emoji winky faces and it's like, no, we need to this isn't real testing. You need to have that, that buy in to maybe sort of carve out part of your media spend that you're willing to lose on more proof of concept creatives. And as I say, get someone in the room that is strong in this area to sort of like lead on creative direction, but also operationally you have to be able to scale these things really, really quickly, which requires some sort of creative framework for you to use, copy and paste and just keep going. Otherwise you'll spend your whole time chasing people in Clickup or slack or wherever, trying to get answers to things. And at that point you're going way too slow to really get the effective learnings that you need.

Thank you so much. It's been a great conversation and I would love to hear from you. What kind of learnings, what kind of insights have you got from this conversation if you can.

My biggest one was the go to Google rate which I actually did not know about. So that's definitely something I'm going to check out as to how meta can help us.

Maybe reveal too much, but yeah, you're welcome. From my side, the thing I love most about these types of conversations is, just being able to sort of openly share this and understand that a lot of the challenges that certainly we have at Personio and I know maybe some of our other people have in the category are also shared by other digital teams in other verticals and other categories as well. So the first thing I do is sort of breathe a sigh of relief that, there's other people out there that are kind of going through these motions and coming up with, solutions to them. I love the idea about, really sort of changing our tech stack as well and looking at different, measurement protocols and adjusting. I think they're the main things that I'm taking from the call.

It's just really great to have these conversations and to just like knowledge share. From my background, I just learned so much just being on calls with both Alex and Rupali. It is interesting just to hear that, the industry is at a kind of a big change point and just how we operate. And a big part of that is this kind of convergence between, how do we shift from performance marketing, which has been geared towards making certain attribution models work to just this kind of transition away from like softer signals and in like creative testing and more kind of creative sides of marketing coming into how we operate to then show up in attribution reporting, which I just think the industry is just a really interesting time right now.

Thank you so much. It's been a great conversation. We will write a report from this. So we will capture the best insights and we will have the opportunity to revise it with you. It's been a great privilege to kind of hear you talk about these knotty issues. It just always impresses me the depth of skill and knowledge and experience in this space. It makes me very excited for the future.

Thanks so much, everyone. It's really good talking to you all.