Automating the Marketing Pipeline | Gil Allouche from Metadata

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Gil Allouche 0:00
The part that I would give myself advice is to reverse engineer to exactly where I’d like to be. And then fill up the missing blanks and so that I know exactly month after month, week after week, what is my projected goal? And how am I doing?

Alexander Ferguson 0:23
Welcome, everyone to UpTech Report. This is our Applied Tech series. UpTech Report is sponsored by TeraLeap. Learn how to leverage the power of video at teraleap.io. Today, I am joined by my guest, Gil Allouche, who is based in San Francisco. He’s the CEO and founder of Metadata.io. Welcome, Gil, good to have you on.

Gil Allouche 0:41
Thank you, nice to be here.

Alexander Ferguson 0:42
Now your product is a platform for pipeline generation. So for those out there, if you’re a CMO, VP of Marketing, specifically VP of demand generation of b2b mid market, this might be an intriguing platform you want to check out now get on your site, you actually have a quote from God are able cog to and I like what he states here, metadata, opera operationalizes marketing data and deploys campaigns in a way that far exceeds the human capability. I find that kind of interesting, far exceeds the human capability. But I’m curious, when you started metadata, what was the problem that you saw, that you set out to solve?

Gil Allouche 1:22
It was very straightforward, my job. And the problem that I needed to solve was how do I generate pipeline for sales, how to generate demand, so that my sales counterpart, back then it was Spotfire, or cueball, it was just, they would be able to take that pipeline and close it into revenue. That was the biggest problem that I wanted to solve, and doing it in a predictable manner. That was the differentiation not doing what they call it a one trick pony. And sometimes successful, sometimes not, but doing it in a predictable manner where the salesperson, she she knows that you’re going to get a discrete dollar pipeline amount that he’s qualified month, month after month, quarter after quarter.

Alexander Ferguson 2:11
It’s that consistency, reliability that you’re trying to do. That’s the differentiation, that it just it just does it for you. Effectively.

Gil Allouche 2:20
Exactly, exactly. It does it for you. And it’s not, there is no excuse behind it, it doesn’t rely on one thing that when that thing disappears, you don’t have it. And I found that the biggest thing, that qualico thing that that if you have it, you have predict the pipeline pipeline. And if you don’t, you don’t have predictable pipeline is the team is the is the team or the human that needs to use your technologies and data and content and creative and channels and audiences and so on so forth. They have this marketing mix. And they they have to the modern marketer has to figure out how to use this marketing mix in the most effective way possible. And that’s what God means when he talks about operational operationalizing data because you can buy the most expensive data that will give you target companies with intent and companies, we’re using your competitor in our competitors technology. And you’ll get a data set that gives you their personal email, corporate email and phone number and in a LinkedIn profile, and you can get the big budget for Facebook ads and LinkedIn ads. But if you don’t experiment and use data driven repetitively, to fine tune your marketing mix into what works, what doesn’t work in terms of end results, the pipeline generated versus any of the vanity metrics in between, like, if you focus on impressions, or clicks or leads or cost per lead or anything in between, that doesn’t lead to eventual pipeline that closes. That’s where the bottleneck is. And that’s what we’re trying to solve.

Alexander Ferguson 3:51
Now, this has been five years that you ago that you started, metadata, I’m sure there’s there’s, it’s, it’s always continually evolving and improving on this. And for those who want to hear more about the journey, definitely stick around for part two of our interview, but to give a taste. If there was one thing that you could say to yourself five years ago that you know, now, five years ago, what wish what would you wish you had known when you begin?

Gil Allouche 4:19
I think reverse engineering. That’s the concept that I would I already had the concept of Pareto rule, right? It’s the 8029 to build just enough and validate it. And we did that. I think the part that I would give myself advice is to reverse engineer to exactly where I’d like to be and then fill up the missing blanks and so that I know exactly month after month, week after week, what is my projected goal and how am I doing in comparison to that because it creates the pattern and the pattern that usually is what leads people to success here in Silicon Valley are fitting a particular pattern that the end without them understanding the investors understand the partners on so forth.

Alexander Ferguson 5:05
not reinventing the wheel, but reverse engineering so that you know where to work back from build it, though, correct? Well, I’m excited to hear more about that journey. But coming back to metadata in itself, the technology, how it’s developed. Tell me, what can you share about the technology that makes it stand out and different from other options out there?

Gil Allouche 5:23
Yeah, so metadata is another technology, we have for USPTO issued patents. And we have the largest proprietary search engine, if you will, for b2b marketers. And so those two components together is what makes the metadata unique. And so the first part is the data set. Imagine that today, when you want to run campaigns, you have to source data, you have to say, Hey, I’m looking for companies who are using these technologies, or from these, the size of company or these job titles, and going after this in your different locations. And it can be very sophisticated criteria, our solution essentially normalized and ingested, most of the b2b data sources out there. So it does the work for you, or first of all, giving you one centralized place where you can create subsets of groups within many data sets at once. You don’t have to go and work separately and do the segmentation. The second part is that we normalized the taxonomy so that we know the difference between an industry on LinkedIn on inside or even referred agent data or embora. So you don’t have to do the compression yourself. So they can really do band diagram. And that allows you to only spend money and time on the right companies and the right people, you never have to worry that you’re Miss targeting targeting the wrong persona or the wrong company. So you’re 100% focused on your total addressable market. That’s one part, the second part of the technology, and that’s our competitive differentiation. That’s what makes us unique, and that’s where most of our ideas around is experimentation, experimentation as a means to an end to achieve your outcome. I didn’t know 10 years ago, when I was in marketer, what campaign to run, I didn’t have that natural instinct of like, I think these words are gonna really hit it strong. And this color of the button, not really my trade my natural traits. But what I did is set up an infrastructure where I can drop all of that all of those possible ideas, and the system will experiment by itself and tell me what stick What did it look this color work with this persona on this channel in generated lower cost per lead is generated higher ACV in your pipeline later. And that experimentation system is our, that’s a big technology behind metadata. So it allows you and that when God talks about in human capabilities, that’s what we’re referring to today, they start to score is doing it manually. And so you have a team of one 510, if it’s a bigger company, maybe they have an agency, they have 50 people, you know, IBM, I think they have 100 people or more working on group and working on their dimension just for particular business units. And so imagine 100, people sitting with huge spreadsheets, a lot of data sets, and constantly analyzing what happened a week ago, two weeks ago, and making some decision and they do the capacity to how many hours they have and how much work you can do. So instead of choosing 1000 possible experiments, they’ll compromise on running 50, because that’s the capacity, a computer. It doesn’t, it doesn’t have to compromise. So you can run all 1000 experiments and quickly, not two weeks ago, quickly say what’s working what’s not and eliminate or quadruple the investment on what’s working. And that part of the technology is what makes marketers our customers successful because they can rely on the system to tell them how to one create a predictable pipeline, that predictable demand for their sales counterpart, but also how to do it in the most economical way. Because it has the ability to

Alexander Ferguson 8:44
do so those two pieces, one data, making use of it and having that housed in there in the right way and being able to segment it. And then the other pieces being able to test effectively. But automatically, all the variations, those two combined is kind of the the core components.

Gil Allouche 9:05
Exactly right. The experimentation is the main is the core one and the data to make sure that the experimentation to begin with is somewhat qualified. So it only goes after the right company to the right people. That’s exactly right.

Alexander Ferguson 9:16
Got it. Now curious on that second part, the the the testing and the automation of that, how much is it truly automated what pieces then does the marketer need or two to play a role in? And what fastest? Does it actually test versus designers create and other people play a role?

Gil Allouche 9:34
I love that question. That’s the that keeps us keeps us honest. So the experimentation indeed doesn’t take over everything. Because not everything a computer today does better and we haven’t built everything yet. And so where we focus on Well, the audience creation is completely automatic. So you can create as many audiences as segmented as you’d like. You know, slice and dice by different criteria, your first party data From your own system, third party data, so on and so forth. So that’s completely automated all the way to find the PII and onboarding those audiences to the different channels. So that part is completely automated, in terms of the variable is the variable that we play with that we, experimentally, we’re talking about the target audience, so the companies and the people within them, we’re going after the channels that Facebook LinkedIn is display, you know, other channels, the campaign type, even within LinkedIn, you can do a conversational ads, it’s like a chat. And you can do sponsored updates. And you can do a regular ad on Facebook to do a cover sale, you can do a video

Alexander Ferguson 10:36
platform is actually managing that those ad platforms, they’re not having to go into those ad platforms, they’re just using their dashboard.

Gil Allouche 10:42
Exactly. And that’s, you know, because we identify the human element or the marketing team, being the bottleneck for execution, if we were to give or recommend, that will still not have the problem, because you still have to do all the heavy lifting. Here, we say, I think we should, you should run this campaign. And he goes and executes the campaign for you. And then he said, I think you should remove the budget and it goes into remove, remove the budget, or lower it or change, changes the creative or changes that word, so on so forth. He does it via a RESTful API into the actual journal.

Alexander Ferguson 11:13
Got it effectively, this is allows folks to be much more efficient on a broad scheme across the entire gamut, to be able to test in each of the channels and not have to manually mess with each one. Exactly,

Gil Allouche 11:28
exactly. And optimize it all, not based on what those channels are trying to optimize for, which is sometimes for impressions or cost per clicks or things like that. Here, you’re often optimizing towards revenue, or the actual KPI. So the longer you let metadata execute and optimize campaigns for you, the further down the funnel, it will optimize towards because the system metric that has access to all of your marketing stack, we doesn’t we don’t go in and say rip and replace HubSpot with metadata instead or change, you don’t have to take the data source. Now equipment that is not at all, you already invested in marketing, automation, and CRM and data sources and channels, we plug into all of that will connect your Salesforce into your Marketo into your boron ag and whatever data you have, if not will complement it to the data set that we have will connect to your Facebook and LinkedIn and execution will happen. And it will happen in the same way that VP of demand will do it it they will update she will not optimize based on impression because it’s on Sunday, she could talk about the CEO or the board members, but you optimize towards pipeline creation or, you know, revenue or things like that.

Alexander Ferguson 12:29
What’s a word of wisdom that you would give to a VP of demand generation in today’s environments, even aside from your product itself, just a word of wisdom in what they have to deal with?

Gil Allouche 12:41
You know, I would say the agile, that’s the most important thing being agile, and many times there is not a fit to metadata, like you said, like if you don’t, there’s particular budget and systems and so forth. But have something I think always is true, always holds water, which is to be agile, not be set in stone, you’re coming into a new position as a as a VP of dimension, the VP of dimension, of course, see what works, what didn’t work, but save a big portion to maybe there are false positives. But sometimes people think Facebook doesn’t work for us because it’s not going to be well, if you apply a little bit of a data set on it. It could work, or chatbots don’t work for me, like, you know, you allow yourself to fail and experiment, experiment, fail, experiment fail. And then one of those, one of those will be magical when and you have to allow the space for this to happen.

Alexander Ferguson 13:31
I like it. Question for you looking forward for metadata, what’s the future? What are you most excited about coming up that you can talk about and want to share?

Gil Allouche 13:42
There are two areas in which product wise for me is where I am very excited. I’m really excited about product. And I’m passionate about marketing. And so for me to see those things coming out and people using them. It’s exciting. So heavy companies like you know, Zoom is in video as a customer or the dreams of the world vendors and seeing those companies starting to adopt the decision tree mechanisms are starting to have more and more things automated adopting more and more automation and AI into their decision making. And, you know, seeing customers winning awards for campaigns that the AI set up. Those things are I’m excited to see, you know, we’re very much in the early adoption stage. And I’m excited to see how we project towards that. You know that through the adoption curve. You know, it relieves the marketing team from their mundane technical repetitive work that no one really likes to do. And it really gives you the ability to work with strategic even within marketing off to work on strategic creative projects, which I think is more more fun.

Alexander Ferguson 14:47
That’s the role of technology. It should be at least to make people’s lives better. They don’t have to do as much manual click buttons, but they can be think creatively and innovatively and he said be agile to testing things. I love it. Well thank you so much for sharing about metadata for those that want to learn more, go to metadata.io. And you can check out for a free demo and learn more about their product. Now, stick around definitely for part two of our discussion though. With Gil we’ll hear more about his founders journey. And UpTech Report answered by Tara Lee, or how to leverage the power video@teraleap.io And we’ll see you guys next time. That concludes the audio version of this episode. To see the original and more visit our UpTech Report YouTube channel. If you know a tech company, we should interview you can nominate them at UpTech report.com. Or if you just prefer to listen. Make sure you’re subscribed to this series on Apple podcasts, Spotify or your favorite podcasting app.