Andy: (00:00)
I think with the current subscription and everything. Let us know if we can do anything. Obviously, I think the last conversation we had for you thought it was quite safe. That’s can be riskier now is this.
Adam: (00:12)
One of our [inaudible] some of the senior people have asked and what per their budget can we optimize the usage of it more for the contributions that we’ve got. So just delays things that we’re now looking at that and where we have, we can’t do it building the stronger business case and how we’re going to increase the usage. So you can see how that impacts us. So yes, [Inaudible] I can’t make any commitments this side of the year, we’re going to have to, it’s going to be something their probably spending into next year, unfortunately.
Andy: (00:47)
Okay. And is Emma still like… I know she was involved a lot last year. She’s still involved?
Adam: (00:53)
Sorry who?
Andy: (00:56)
Emma,
(00:56)
Emma, No she’s on holiday at the moment, but she would be involved in this process at all. It’s mainly myself, Matt, and the new business team who I’m working with who contribute budget as well.
Andy: (01:14)
Okay.
(01:14)
Perfect. That’s helpful. [Inaudible] And just the last one from me Andy then Adam. nd then Adam. When [Inaudible] was chasing me and from OMD crate and Thermine Grungy, who reports to Matt, In fact right? About her long can I access. And Is she involved in those conversations? Is there anything I can do to just say hey? And have you let [Inaudible] know that she can contribute budget. And then she wouldn’t have the say on that, but just to make sure communicated to you and to her.
Adam: (01:44)
[inaudible] There is last day their isn’t Friday’s. So we’re actually…
Speaker 3: (01:47)
[Inaudible]
Adam: (01:49)
[Inaudible] So, just stay best leave [Inaudible] and Matt, who analyzing my use of budget.
(01:55)
Fine.
Adam: (01:55)
and our uses.
Andy: (01:59)
Cool. all right well Adam, I can introduce you quickly, But be great if you did as well. Just Andy for your background, working with Adam now for about three to four years. And he has been running like social analytics team, social intelligence team in OMD. And servicing different clients of theirs with social media data, including our online video data. Adam, anything else you want to give background for Andy?
Adam: (02:29)
Yeah, I mean he served two years there a lot online statement clients like [Inaudible] disney, And just a lot of trend reporting and videos. Sort of understanding what UGC works for parents and kids.So Disney channels can create better content, looking how we can do it for our clients in entertainment as well. But gaming as well. Activision was quite a big important client for us as well as your wafer. So sports and some typical content as well. Looking at building our trends they’re offering in there. And the develops that we found quite interesting from a tube perspective in the past years the also categorization. So the type, the taxonomy that you’ve built around topics, Stuff that can help us do this at scale easily [inaudible] it’s very important because we’re looking at vastly billions amount of the views data. So being able to create trends tracking systems is really interesting for us, I think. But in terms of how we use tubular data trends is probably the biggest use case for us.
Andy: (03:46)
Now it makes sense, That’s great feedback for me. Let me share my screen one second. I almost got ran over literally at the front, walking into the office. So a couple of minutes late. Alright, So here’s my presentation. So my objective here, I lead new product development for our buy-side customer’s agencies and brands. Little bit about me, I used to work at a DSP on the ad tech side called Quantcast. I used to build the ad targeting products over there. I moved over to tubular about nine months ago, really loved the data sets, and what was possible there. And so I’m building some new products specifically for agencies and brands. I’ve been in the product development cycle of what I’m about to show you. This Q3, where I was really validating the value sort of proposition and what pain points that customers are having. Sort of zero down into a core one or two that I wanted to focus on and in this quarter is holding about designing the user experience of what I’m going to show you. It’s a look and feel on how it works. And so my objective is to demo the tool and then just have an honest conversation with you about hey as a agency, These are the features that we would value, These are the ones that we would actually pay for, These ones we wouldn’t pay for and just have an honest conversation about, so we can know hey, which features need to go into the MVP what pricing levels, should we expect from customers are you willing to pay things like that. Did that sound [Inaudible]?
Adam: (05:26)
Yes, Sounds good.
Andy: (05:29)
Let’s get straight into it. So naming the project at the moment create ROI, This is not going to stick. Every product did have a name has never ended up that name. So the name sucks.
(05:40)
Wound just call it that though.
Adam: (05:42)
I really wanted to focus on the media buying process. So, tubular we have access to this amazing data. However, we’re not by contract allowed to optimize or execute any media. So that made it easy for me I’m like, all right, let’s focus on research and planning phases. Research comes planning, media planning, maybe even media buying. But we can all play in the execution and optimization space like in cookies and whatnot. So I just wanted to ground you in where we’re playing. Essentially, the problem statement for this is using views or followers can be a misleading vanity metric used to determine who are the right content creators to partner with. So what this product actually is, is it ranks which creatives are attracting an audience with an elevated propensity to purchase any product category or brand on Amazon. How we do this, we can actually tie a creator’s video content consumption to that audience’s Amazon purchases within a defined attribution window. And the hypothesis is that you can actually increase the ROI of your media or influencer spend by identifying creators that demonstrate the highest purchase rates. Does that make sense?
Adam: (07:02)
Yeah, that’s really interesting in terms of how we integrate this on to our E commerce team. Because they’re always asking us how do we actually segment influences who are number one, probably can authentically promote product but also drive that and propensity as well. So linking it to Amazon’s really, or Amazon data is really interesting.
Andy: (07:29)
I really wanted to focus on business outcomes. Let’s focus on sales rather than views essentially.
Adam: (07:37)
Yeah.
Andy: (07:37)
So I want to tell you a little bit about the methodology. So obviously, you can’t cookie on Amazon, you can’t cookie on Facebook. So we actually use a panel. And where we can see everything that that panel does, the panel is about 20 to 25 million unique devices globally. It’s a behavioral panel. So it’s action based, Not survey based in the US I always pick on MRI. I think the UK version is TGI.
Adam: (08:03)
Yeah,
Andy: (08:05)
And the core actions that we get access to would be what are they searching for on Amazon? What’s the age, gender, and geo of where they bought and who they are? And then, the specific we can see what product they viewed, what the reviews they read, what they added to cart, what they started to check out, and what they actually purchased. So you can see that full funnel from search through view-through conversion. The data set covers Amazon purchases on desktop and mobile web. It’s not currently mobile app. We are planning to expand that we have identified a second-panel provider that can give us that, but the initial will be just desktop and mobile web. Another thing is data is refreshed daily. And such of significant difference to like MRI in the US, so I compare the panel size, so it’s about a 1000 times larger. Then MRI It’s behavioral rather than survey-based, which you can get into the problems of like I want to answer aspirationally. And then the refresh rate is daily rather than once every six months. So we believe that this is actually probably a step function improvement and accuracy and benefit. And the other thing is we’re starting with Amazon, but it’s really easy for us to add other purchase sources like Walmart, Target, Costco, Sephora, GameStop, you mentioned gaming. Also, you’re very you case,, specific ones, Marks and Spencers, Selfridges, boots, things like that. So this is a little bit about the data set. And then what are we solving for? So obviously, this is a brand new product. This has never been done before. So we have to be very careful about what we try and solve and what we don’t, so this is decision support for identification comparison selection of influences, publishers, and back brands. We can do this by ranking which creators are and attracting an audience with an elevated propensity to purchase any product. So I need to like let that sink in. You could look at competitive brands. You can get the access to who is buying your competitive products. What we’re not solving for in this first version, we are not looking at causation of individual videos, driving specific product sales. We’re not trying to say this video drive pro ten extra incremental sales. What we are trying to say is this creates audience bought $10 million worth of this product. Also, we’re not trying to solve for influencer campaign reporting at this stage, Really just the identification, comparison, and selection. And also, we’re not solving for creators’ brand suitability. That’s really on the brand to determine which creators content sort of fits with their brand values and whatnot. We’re just providing the data and the numbers. So let me get into…
Adam: (10:51)
I guess the interesting thing would then to be, you need to split test this so you need to look at have they people have used your creative ROI product on a same campaign versus not using it or having a [Inaudible]
Andy: (11:06)
We’re not trying to go for causation. If I was trying to look at causation, we would look at do some sort of control and exposed Hey, what’s the baseline? What’s the baseline purchase rate? And then for this group that saw an ad, what was the purchase rate and what’s the lift? We’re not trying to solve that that’s a really difficult problem that we could solve in the future. But we want to go for something a little simpler and just look at the overall creator’s suitability for driving specific sales.
Adam: (11:34)
Say I mean, I get that causation of individual videos, but if it’s called ROI and you want to say, use this tech to find influencers, you have a pensity to drive purchase, then doing it versus Monza how you do it normally having measuring of effects of that would prove the ROI.
speaker 1: (12:00)
That why I said the naming sucks.
Adam: (12:04)
Maybe we call it [Inaudible] could be the naming.
Andy: (12:09)
Let me get into the prototype. I’m going to go through fairly quickly. I’m going to do a gut guided walkthrough. So you start with the end in mind, What’s the product that I’m interested in looking at, And so this is a clickable prototype. Obviously, it’s not connected to a system behind it. So there’s two ways you could try and find the product, You can click through the hierarchy, so if I was looking at makeup, I’d go through beauty and personal care to makeup to eyeliners. And then I could choose all eyeliners, or I could choose a specific individual skew level product eyeliners, So we built this experience for benefit cosmetics. And so I’ve gone down to the individual skill level product, I can overlay some filters like it has the brand new overlaid. If I was one level up at all, all eyeliners I could say Hey, show me all eyeliners that will benefit cosmetics. Obviously, we’ve gone down to a skew level product so it has to be benefit cosmetics obviously, I could also filter my creator results by the number of followers and subscribers, potentially that Geo-location Hey only show me influencers in Italy or wherever it might be. And then the creator type so a creator could be either a publisher, like a Buzzfeed, It could be a influencer like sir Alex Ferguson, or it could be a brand like Yo, Activision, for instance. In this example, I’m going to actually go to the level above this. I’m going to go to all eyeliners also a built-in the ability to select what your attribution window is. So within how many days of the audience watching the video did they purchase? so, you could select it at one day. You could select it at seven days, 14 days, 30 days, I’ve mimicked what the video ad serving common attribution windows are here. For the example of this, I’m just going to leave it at seven days, and it’s looking at the last three months worth of data. And so there’s some core features. So we’re looking at all benefit cosmetic eyeliners that were purchased within seven days of the person watching the video. And so the first core feature would be top-line sales numbers. So over the three month period, $12.5 Million worth of benefit cosmetic eyeliners were sold on Amazon. That was the total purchases at 1.1 million at the average selling price of $11-22. So top-line sales numbers is feature one. Feature two is around insights of who actually purchased the products. So you have gender split male, female, you would have age and agenda split, So 18 to 24 males and females, we can show that as a composition, but we can also show that as an index. So how does that index versus the UK average for those demographics, And then finally, the Geo-location. Hey, where did these people actually purchase from? So that’s feature number two is demographics. Feature number three is the actual creator ranking list. And so in this case, there’s some new metrics available in the system. So it is sorted by the hero column. The hero column is the one directly to the left here, and it’s using a tornado graph. And so it is sorting based on who has the highest numbers for that particular column. In this case, it’s total sales. You can also see average selling price, And I can actually go to the right and then scroll to the right and see other things like unique view was reached sales per unique plus sales per minute watched of content. I could also… Customized columns, Hey, I want to look at YouTube specifically or Facebook specifically or change around the ordering of my metrics and stuff like that. I’m not going to get into that. The interesting thing is, depending on what your objective is, is going to change what the right metric that you want to use. Total sales could be a little misleading because it’s a function of how many people does that person have in their audience, So a mega influencer is always going to have or is likely to always have really, really high sales numbers because they just have so many people in their audience and that audience buys stuff online. So I would actually use if I was trying to evaluate creatives based on their audiences purchasing behaviors. I’d actually use sales per unique viewer. And so someone that has a much higher sales per unique view, that’s saying that the average person in their stable or in their audience by as much more of benefit cosmetic eyeliners than another.
Adam: (17:16)
And when you’re thinking about ROIs, well you’re paying them to do less content to get more results. So…
speaker 1: (17:22)
Yeah,
Adam: (17:22)
One video here, which might be… that’s quite nice.
Andy: (17:26)
You could also do things like, Hey, let me look at the average selling price for this same product. Conde Nast is getting $79 on average, for the price point, Whereas, Michelle Phan is only getting $24. So you could also bucket people based on their audiences, low end, medium, and high end purchasing behaviors. So this is core feature Three. Obviously, you can customize the columns, We’ve got it set with conditional formatting, so it makes it really easy to sort of look across and see Hey, is this creator consistently strong across all the metrics and they’re in green and blue or they consistently Pull and read across the board, so it makes it easy to see if they’re doing well consistently.
Speaker 2: (18:14)
Want to know, and I’m sorry because, on the previous slides, you mentioned that it’s [Inaudible] purchase any products, but we’ve selected a particular product category type. So yes, quest Why would it be any product if we’ve selected a product type?
Speaker 1: (18:32)
So, if you have a client that is in this makeup, you have benefit Cosmetics, they’re going to care about this. However, if I had toys, I was actually demoing this to Mattel yesterday, and like, Hey, we’d love to do it for like Nerf guns. So I changed and clicked through this I’d actually go through like toys and games and click down to that product that they care about. Maybe overlaying Hasbro or maybe overlaying a competitive if I wanted to look at their competitive sale. And so choose your own adventure. What level of granularity do you want to get to?
speaker 1: (19:07)
And do you want to look at my brand, d you want to look at someone else’s brand, So the flexibilities in the selection piece, up the top though, [inaudible].
Adam: (19:14)That’s enough granularity, I think I was more confused that I said any product of women, any product that they can get some [inaudible].
Speaker 1: (19:23)
Any product that they can sell on Amazon, essentially. So, all right, I’ve got my ranking. I want to actually start, measuring and like, Hey, I want to start building a list together. And maybe I already have a list that I want to see how they compare, versus, what their actual sales are doing. So I’ve selected Toxie I’ve selected Jeffree star dope 2111. James Charles, I’m going to save this. And then I’m going to name this, top four trending creators for eyeliners. And then now I get a new screen where I can actually measure over time and see the historical sales figures for that creator and their audience for that specific product selection, So at the moment, it’s total sales. I could actually change this to Hey. It’ll show me what the average sales per minute watched is. And then I can see the top T is doing well in total sales, but also total, minutes watched, Make it easy to… Sorry, I am going to go back for a second… Then do things like hey, I want to add a new creator list add Michelle Phan to this. Let’s apply her, there. And then she’s now someone I’m tracking so you can be flexibly dropping people or adding people over time. So this is the historical way any creator can a one month bump, but you want to see that your creative audience is actually driving sales consistently over time. I heard in discovery, talking to Nike and talking to Adidas and all the people that went into helping develop this. Influence doesn’t stay static over time. It tends to go up or tends to go down and so this historical sales ranking will give you the ability to see that and also annotate it and say, ‘Hey, we started an influence campaign right here, let’s watch what happens to average sales when they start talking about this product.’ And so that –
(22:27)
Adam.
(22:27)
Going on [inaudible].
(22:29)
On these because a lot of the influences are showing here, very top tier if he went to like the nano micro influences of a covered as well or would it be only mainly the ones that you can actually get enough data for.
Andy: (22:46)
So micro yes, definitely. And when you get down to the nano level, That might be more challenging. But actually, let’s switch gears. So that the demo So that you’ve got four key things to this that which product or combination of products do I want to select? There’s the top-line sales numbers, the demographics, the creator ranking, and the historical sales for that to create or set. I’m going to change gears now, And-
Adam: (23:15)
Andy.
Andy: (23:17)
Yeah.
Adam: (23:17)
Just one quick thing, what you’re showing there, I guess Adam and teamwork a lot with brands who are doing owned and operated company themselves versus influencers. So Create content, Would we be able to do that as well?
Andy: (23:32)
Yes.So publishers are part of the creative set. Actually, I got to go back to the pitchers for a second.
Adam: (23:38)
I can see more brands. So like Hasbro could be put in there.
Andy: (23:44)
Absolutely. So let’s talk about the core use cases. So core use cases for this product would be identifying and evaluating influences for influencer marketing programs, very simple. But if you looked at the publishers specifically, you could use it to compare, and select publishers like a Buzzfeed versus a Conde Nast for direct site buys, for PMPs, for customer publisher integrations, It can tell you, who is doing what. And really interestingly, I was looking at the underlying data for this. We got a month sample from our panel provider, and we could actually see that Conde Nast had a two X sales per minute watched of content compared to Buzzfeed for beauty, which was… I was expecting that because Conde Nast is a big beauty publisher. But it was just nice to see the data on that. And then the third one of the identifying other brands with heavy purchasing audiences for co-marketing opportunities like Roma and Tide laundry [Inaudible 23:30] something where they may not exactly be competing for the same customer, but they had error Jason and then the last one to become questing, So identifying and trying to appropriate who are driving the highest sales of competitive products. So those would be the core use cases. And then depending on what your objective is, there be best practices. So if I would get total sales, that’s really telling me which creators audiences are buying just the most in a total of a certain product or category. However, if you are looking at average selling price, that’s really which creators audiences are in the low, medium, and high end of a product category. Sales per unique, that’s probably the one I would use, which creators audiences are buying the most per person they’re actually reaching. So that’s the apples to apples comparison. And then probably more for the publishes sales per minute watch which creators have the highest purchase rates based on the amount of time spent viewing their content. So these would be like four of the new metrics that would be going into these, this UX. And so I’m going to change gears now. I don’t believe in just ramming pricing down a customer’s throat. I actually want to understand what features you would actually value and which ones that you would be willing to pay for as a company. And so the first question, so what now at this stage, we’re going to start the R &D and technical actual development of this in Q1. So in January, we’re starting, and I need to know what goes into the MVP, essentially, it’s the first question I have, of the four components that I described in the solution, I want you to allocate some points for them. Essentially, zero would be, it’s not really valued to us, One would be limited importance, Two would be a nice to have, and three would be a must-have. You only get five points, So there is scarcity in that. So that means you could only have one must-have in one limited Value, sorry, And one nice to have. You could have one must have and two limited values, however, you would for agency, I’m really trying to understand which features you would value and why. So could you walk through this and apply five points in total across these four different sorts of features that would potentially going into the MVP.
Adam: (25:55)okay demographics in terms of audience. So I’d put that as two. I’d put your estimated total sales as zero, as long as we got the creative historical sales performance as a three. And because if I’m getting that, then I can get the total.
Andy: (26:18)
And so you’d have to rank create a sales ranking at zero or two.
Adam: (26:22)
Well [Inaudible] because I’ve been in mean historical performance is really important, and that you could get the total from that. So if you’re going to look at the ranking, I’m less concerned about because You’d have to rank them against how much you’re spending. So how much you paid that influencer. So you need that baseline data of, because you could be spending hundreds of thousands of micro-influencers or on top talents to actually create that ROI to rank them. So I think the ranking it’s [Inaudible] would, without having enough spend data should probably sit and, it will be up to the Agency to do that ranking really so when [Crosstalk] this influencer costs us X, therefore, against the historical sales performance is X.
Andy: (27:09)
Exellent.
Adam: (27:10)
So I’ll put it as that really.
Andy: (27:15)
And then another thing that I’m sort of working through is this challenge of where… because it does give you flexibility to what level of the hierarchy do I want to actually look at. So let me rank these real quickly for you to make it easier, So essentially, you can go to the category level, So all makeup if we chose like toys, maybe it’s like I don’t know, old bored games maybe or something like that.so you could go to the old makeup, old brands, You can look at old makeup, a single competitive brand. So I found benefit Revlon or Maybelline or something like that, or I could look at old makeup, my brand, benefit cosmetics. I could go one level deeper for like makeup to all eyeliners and same thing all brands, my brand, or a competitive brand. Or it could get down to the individual single product level. So benefit cosmetics roller liner makeup, three-ounce bottle that in black or a competitive product. Revlon Matte eyeliner, same ounces in black. Could you rank which do you think levels of hierarchy and combinations like from most important to least important of which you would actually use here? from one to…
Adam: (28:33)
[Inaudible] well, the questions we get from clients are going to be growing granularity. Say if you could go to single product level, and then that’s amazing. Getting to … Depends on what days you’re getting. So you might have limited data. So you have to look at a brand level or a category level, But I would definitely push anything that’s product up top because that granularity will be really the most important.
Andy: (29:00)
And yours first and competitive second?
Adam: (29:04)
Yeah, own the product first and then competitor second.
Andy: (29:08)
So you want to go from most granular so the single products to the least granular so I guess subcategory would be second?
Adam: (29:16)Yeah. So your own product, then competitive product, then [inaudible], Then it goes into the brand stuff.
Andy: (29:25)
And so the same thing. My brands,
Adam: (29:27)
Yeah.
Andy: (29:28)
…Someone else’s brand or brands.
Adam: (29:30)
Yeah. That’s it.
Andy: (29:31)
So this is a good kind of how we would rank it from most to least?
Adam: (29:36)
Yeah. I imagine it’s the most challenging for you to build as well.
Andy: (29:39)
Yeah. Obviously, with single products, if you’re in the long tail, there’s probably not enough data to get statistically significant data, but there will absolutely be, medium and high volume products that we can absolutely get down to the individual product level, which is exciting.
Adam: (29:53)
Yeah.
Andy: (29:54)
Also, talk about rank the most important purchase attribution windows. We might not give the option to do all four of them, to begin with. So what would be the most important to the least important in terms of the different time from video view to purchase essentially? [Inaudible 30:14]
Adam: (30:15)
And because you’re doing it within social. An influencer, you would say one day because of the consideration, that using an Instagram creator should be fairly quick transaction, One day I’ll do 171430. So that I think, depending on the campaign, but yeah, probably put it more [crosstalk 30:37] more of that-
Andy: (30:38)
So most granular to least, So the next one would be, Hey, I’ve presented three use cases we can rank influences, we can rank publishes, we can rank brands, Can you go through and say which of these would be the… Where they would fit in this no low, medium, high value. And you don’t have like a scarcity here. You can rank them all high. You could rank them all low, where would you say each of these,
Adam: (31:04)
Influences high publishes high. And then if you’re someone like [Inaudible 31:13] for Disney you’re trying to value their brand or stay for partners, it’s still very high for us, Be quite useful, interesting to know.
Andy: (31:22)
Okay. So I talked to Mattel literally yesterday and their brand, and so I think the business model for them is going to be different to an agency, obviously. And so I want to try and just have a discussion about what are the direct and indirect ways this creates value for your business being an agency and are you able to quantify what they might look like. And so for example, headcount efficiency, if you’re a brand sales lift, maybe negotiations and having data, can you talk about the different ways that you could see us adding value to the agency?
Adam: (31:59)
And I think the negotiation aspects interesting when we’ve creators, Or at least for me more on having better management on the effectiveness of an influencer. And so we know we’re negotiating with the right influencer, and the sales list as well. And, I mean this are I .e.s
Andy: (32:33)
So that one’s interesting. I’ve been thinking about it for an agency. So while it’s, you don’t get, obviously the benefit of the sales lift, I guess if you are outperforming other agencies in terms of the sales lift, you can drive your customer that benefits you indirectly, I guess.
Adam: (32:47)
Yeah, exactly. We’re always looking to prove the ROI as an agency. I know it influences the selection. So I’ll put that in. The other ways allows us as a business into show better integration with e-Commerce, So you especially an agency to be able to have different show how we can integrate intelligence teams, influencer content teams, and e-Commerce teams in a unified influencer proposition. And [Inaudible] often they’re very separate, so you’ll get e-Commerce teams looking to use influencers. You just now handle products and can’t do very authentic [Inaudible 33:34]. Then you have content teams who just, who choose creative influences. You don’t have to peddle products. So, this will help bridge the gap, and then you have intelligence teams to tell you, always make the right decisions. So say this helps bridge those gaps with some data.
Andy: (33:57)
Excellent. Any, anything else, [Inaudible 34:00]?
Adam: (34:00)
That it, can say that.
Andy: (34:02)
All right. I’ve got a list, like a value waterfall, and I just want to try and quantify some of those things that we chatted about. How many people at OMD would you say how many headcounts are performing these functions of identification, evaluation, and comparison sort of selection of influences or publisher brands that we’ve spoken about.
Adam: (34:25)
Headcount within our agency?
Andy: (34:28)
Yeah.
Adam: (34:29)
15 people would be using it maybe.
Andy: (34:36)
And how much of their time on average per week would you say that they spend doing this specifically?
Adam: (34:42)
Well selecting influencers, vetting influencers.
Andy: (34:57)
Mm-hmm [affirmative] 50% time, 20%, 100%.
Adam: (35:02)
It’s something you’d it’d be so difficult to answer really influencer specialist team. Let’s say somebody who was an influencer FTE and identification aspect would be at least 20% of their time.
Adam: (35:20)
So for specialists, yes, 20%. But I think for everyone else, it would probably be under 2%.
Andy: (35:29)
Okay. So let’s put maybe 10%, or 20%, like eight hours a week.?
Adam: (35:35)
Yeah. You can put it that.
Andy: (35:38)
How much efficiency improvement do you think this tool would give an employee who’s doing this? Do you think it would speed them up 50%? do you think it would be a 25% improvement? How much do you think it could minimize the time spent doing that?
Adam: (35:55)
So the ROI is something that…. So there are quite different ways you select to influencer. Are they [Crosstalk 36:03] driving sales, but also the other measures, so how are they creating an engaging content? Are they relevant? Is it best in class content? Do they drive high engagement rates? Are the right personality fit for my brands, all those other things, which you look for. What this would do is maybe help create the shortlist of people you could lookup. And so you’d have to probably look at its stage first. The first question you ask is, are they going to deliver sales uplift? Once you’ve done that, you can then look at the difference, create a shortlist, and then you go deeper. They are the right creative fit for my brand, and all that so, I think it might help with the first phase of getting to a shortlist. So it’s quite [Inaudible 36:52] a bit of time on that. Or…
Andy: (36:54)
I guess it’s not always time that it saves, it might just redistribute where they’re focusing their attention. So Hey, I’ve gotten to a shortlist of people that we’re certain they’re going to drive sales, now let’s like get deeper and more thorough in being able to really evaluate each one of those people that hit our shortlist. Maybe.
Adam: (37:12)
Yeah, I guess it was to gets rid of the guesswork with a client as well, who often have.
Andy: (37:18)
Could you estimate some sort of efficiency improvement for this waterfall here?
Adam: (37:24)
You want your percent figure in terms of time spent or?
Andy: (37:27)
Yeah, well time spent, but we could also think of impact as maybe efficiency, like some sort of improvement that you think that this tool would give the actual people cost of your business?
Adam: (37:39)
Well, if we take that out, you would have to be on time spent because that’s where we place efficiency. So you take that and let’s say it reduces the time of vetting influencers by 50%.
Andy: (37:54)
Okay. Would this eliminate any legacy tools that you use?
Adam: (38:01)
No, it integrates supplement because it’s not the answer is part of the answer.
Andy: (38:07)
Okay. The next one isn’t really like something I’m looking for something. It’s more like Hey, influencer publisher efficiency, like getting the same output at a cheaper cost or a larger output at the same cost. This question was more for like brands, but like what would be the impact to a business if you could lift sales by 1% or 5% or 25% or 100% essentially. And obviously, someone that has a really legacy of 50-year-old product, a 1% lift for them could be amazing. Brand new product you can absolutely see a 100% lift so, is there any way we can quantify the impact to business based on the lift that it could potentially give a customer?
Adam: (38:48)
So we’ve done beacon footfall traffic, influencer re-targeting before, where we look at sales offline, so targeting people with content who’ve got certain footfall of the store, in terms of the percent of that striven, it would be in the higher figures, it would be around the between the 25 and 100. Digital is quite the more difficult for me, to… Because I’ve really done this sort of thing, but you would estimate it’s likely serve to [inaudible 39:28] in your 25%.
Andy: (39:28)
And what would I put here?
Adam: (39:30)
You can put it in your 25%.
Andy: (39:33)
Okay. But I was thinking more about the agency potential use cases would something like this help your pitch win rate, or pitch deal size. Could you see it impacting these two?
Adam: (39:48)
If it was a specific, influencer based pitch or a client that’s got a high influencer focus, i.e. In the beauty category and there, and it was all about selling product, you could see it having a 10% likelihood increase.
Andy: (40:10)
And what about deal size, do you think are they fixed thing?
Adam: (40:15)
This is more about making more, effective decisions or influencers. So, getting better bang for your buck, So the problem is the clients come to us, they’re spending too much at the moment, although I am as well. So at the moment, it depends on it just getting more bang for that current budget or rather than using it as a tool to increase budget and they are reducing that channel planning tools for that, to make a record.
Andy: (40:48)
Okay,
Adam: (40:49)
This is more about making like a better, I wouldn’t say it’s about increasing your influencer of budget unless you’re proving the value of influencers against other channels, which is not doing it’s proving the value of creators against other creators. I wouldn’t say we would increase the pitch deal size.
Andy: (41:10)
No problem. What about the last one the negotiation piece? Like if you go to the table, being able to say, Hey, influencer or Hey publisher, we have data that says that these guys are driving more proposal, how much better do you think you’d be able to do in negotiations with that data?
Adam: (41:29)
With influencers I think… What kind of figure you’re looking for another percent here?
Andy: (41:40)
Like maybe percentage. Maybe we’d be able to get 5% or 10% or whatever better rates?
Adam: (41:45)
Yeah. I’ll keep this a 10%.
Andy: (41:48)
Excellent okay. So are you had to are there any alternative solutions that you’re using today to get similar information ,and if there are, what generally do they cost?
Adam: (42:04)
So we have access to TGI ,so we do get data from there, but we don’t link it to the Amazon data or your insurance database.
Andy: (42:14)
Mm-hmm
Adam: (42:14)
…We probably do get data from Amazon, I imagine as well. That currently link that too influences though. In terms of what we use for influencers, you won’t get stuff which rich data. I haven’t seen alternative solution which does this in terms of looking at that sales attribution. It’s on your own campaign that you’ve run.
Andy: (42:40)
And you suddenly can’t get it for another brand. It’s like a competitive print.
Adam: (42:45)
No. So [inaudible 42:47] but what you do get, we did use a tool called speaker, which estimated how much you should be spending on an influencer, [crosstalk 42:58] exist anymore. I’ve tried to give you the reasoning about gives you an estimate spend or create certain types of influences. You’re just spending that much on them. That’s the cases we’ve had.
Andy: (43:11)And so, thinking about the actual price point, if we were to sell this, like a SAS product, not like seats and giving you access to everything but pricing it on like, Hey single product, as I’d shown you a product level, looking at one product at a time and charging you a one product at a time, what do you think a reasonable price per product measured, would be? so all benefit cosmetic eyeliners would be one roller liner, another one would be two, what do you think a reasonable price per product measured could be?
Adam: (43:48)
So this is every time you want to look at it and [Inaudible 43:50] like a token, you would [Inaudible 43:52] search for product [Crosstalk 43:53].
Andy: (43:53)
You get a, I guess like ongoing historical numbers for that, you can keep coming back and it’s like your ongoing looking at the old makeup and then maybe for like gaming, you’re looking at certain gaming ones and that would be two, And if you’re looking at Disney, that’d be three or, and so someone like Unilever might have tens of thousands of products, whereas that small business might have one product, whereas an agency that has 20 clients might have, 20, 30, 40 products that they might be wanting to measure-
Adam: (44:25)
$100 maybe per products, If that’s a search.
Andy: (44:31)
What would you consider to be an expensive price per product measured?
Adam: (44:37)
Any, so if this is an agency using anything above that would be expensive.
Andy: (44:51)
Okay.
Adam: (44:53)
There you go. That’s my answer.
Andy: (44:54)
What would be a cost prohibitive product? Price per product measured?