Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Thoughtful Edge, Episode 5: Interview with Igor Ivitskiy world class export of the Google ads
48 Plays1 year ago
In this episode, I talk to Igor Ivitskiy one of the leading world export of Google ads. The conversation concerns digital advertising, including trends, budgeting, marketing automation, AI, and data privacy. We address how to incorporate new technologies and stay up-to-date with the latest developments, as well as the potential risks and rewards of AI in advertising. And also discuss the future of digital advertising and how the industry will likely evolve in the coming years. #leadership #management #careerdevelopment #thoughtfuledge #google #googleads #adtech
Recommended
Transcript
Introduction to Thoughtful Edge Podcast
00:00:05
Speaker
Hello, hello. This is the Thoughtful Edge podcast, and this is the episode number five. I'm Dmitry Savchanka, the host, and today we are going to talk about the Google Ads. Since we are working in the ad tech space and
00:00:21
Speaker
working with the marketing and technologies related to that.
Meet Igor Ivitsky, Google Ads Expert
00:00:25
Speaker
Today is my guest, Igor Ivitsky, one of the leading world experts in the Google Ads. And we are going to talk about those amazing technologies and the amazing space which Google Ads offers us. So, Igor, could you please...
00:00:49
Speaker
Sorry, sorry. You can call me Google. OK, yeah, makes sense. Makes total sense, as we already mentioned, like you're an expert and you're almost Google.
Igor's Background and Advertising Experience
00:01:01
Speaker
So nevertheless, Igor, could you please introduce yourself, tell us about your expertise and about your experience?
00:01:07
Speaker
yes hello to all so i'm from ukraine basically and a year ago i relocated to the united kingdom i am a owner of the most significant google ads online school in ukrainian and russian segment of the internet and recently i started my new i would say new path in my life in the united kingdom in
00:01:36
Speaker
English language market. And now I work with not many ads clients, but with pretty big clients. So some total amount of spending on Google ads of my clients is near 13, 14 million dollars per year. Sorry, per month, not per
Trends and Challenges in Digital Advertising
00:02:05
Speaker
year.
00:02:05
Speaker
13, 14 million dollars per month.
00:02:10
Speaker
Yeah, and my background, I'm a PhD in math, in math modeling, and I'm a former professor, a young scientist of a year in Ukraine, and my scientific backgrounds give me the opportunity to look deeper into Google Ads algorithms and to understand at the more ground level how it actually works.
00:02:39
Speaker
Wow, wow. Yeah, that's impressive. Those multi-million budgets, they always something that one can perceive as a hard thing to manage, but you could really crack that and understood how to better do that. And I have a first question for you.
00:03:02
Speaker
What are some biggest trends you're seeing in the digital advertising right now and particularly with the respect of the big media budgets because like for my experience working at kines and with the media agencies i see that.
00:03:17
Speaker
Like to manage the big budget, we just need more people. We just need to scale respectively, hire a huge team and have all those processes for distributing or managing them successfully. So what are your thoughts on that?
Role of Automation and Algorithms in Ad Campaigns
00:03:38
Speaker
So in general, the biggest trends that I can see, first of all, is using the maximum amount of data that we can collect. So now, like the rise of geolocation services, micro-targeting, for example, in one project, we use system-based on open Wi-Fi access point, so we know where
00:04:08
Speaker
exactly locate every Wi-Fi point at this point and we can know like just like general example if somebody connect to Wi-Fi near this place so very we can expect that this human visit to for example Selfridges shop and
00:04:39
Speaker
In future, we can show to him or to her, to this person, advertising of another luxury brands, for example. Also, it's interesting that you mentioned that when we speak about big budgets, we need to scale our team as well. But on my experience, when I worked with
00:05:06
Speaker
10, even not 10, even 100 times less budgets five years ago. So five years ago, I worked with budgets like $200,000 per year, per month, and so on. And I had a much bigger team than now.
00:05:30
Speaker
because a lot of work that previously did by human now we can
00:05:38
Speaker
do by not artificial intelligence but by math models by algorithms for example one of the like most I would say hard work in search advertising is like to investigate keywords search terms queries find negative keywords and so on and previously my team
00:06:06
Speaker
spend a lot of time just to
00:06:10
Speaker
just to look at thousands of search terms and find some negatives from this incredible big list. And there was always a place for human mistake. For example, one of the biggest mistake in my memory, it was not my team, but I was an auditor of this account.
00:06:36
Speaker
And the company, it was producer of the unmanaged hierarchical vehicle for agricultural, military, and so on, so pretty big company. And the person who checks these keywords, he doesn't, didn't know that the main keyword UAC abbreviation. It's abbreviation from unmanaged hierarchical vehicle. And he added
00:07:06
Speaker
word unmanet to negative list like unmanet what does it mean i don't know what what does it mean unmanet i will add it to negative list and after this they felt several months like
00:07:21
Speaker
very incredible drop in productivity and efficiency and they invited me to check the problem and I found out this problem and this very good explanation how we can have big problems in overall strategy just because of small mistakes of junior
00:07:44
Speaker
position specialist, but now I use a lot of automatization, a lot of scripts to automate this process.
Scientific Methods and AI in Ad Investment
00:07:55
Speaker
For example, I think last...
00:08:00
Speaker
one year I even didn't pay attention to search I pay attention to search terms in search ads for big budgets but usually I use another approach when I collect all search terms create engram engram is like I
00:08:24
Speaker
separate each words to n-gram and now at the second step I calculate by the math module like how much money this n-gram
00:08:40
Speaker
will give me if it will consist in several keywords in several search terms. And after this, I can, like with math precise, determine exactly which negative words I should exclude. And for my surprise, the very famous, you know, negative words like free, download,
00:09:11
Speaker
very often not worth to exclude because even people who looks at something free can pay and pay a very huge amount of money, especially if we speak about B2C segment.
00:09:30
Speaker
Nice, nice. Now this is very impressive and especially those parts when like you just leverage technology to automate processes and do some stuff less error prone like avoiding mistakes and ensuring that everything goes as planned. And you also mentioned that you are using some scripts. Are you doing those scripts yourself or
00:09:55
Speaker
How do you do that? Like for me as an engineer, like it's a very interesting question. I can do it. Definitely. I'm former programmer. I'm former programming informatics teacher in the university. I'm professor of computer sciences. So I can do it, of course. But for me, several
00:10:18
Speaker
Yeah, and previously I wrote these scripts by my phone, but I realized several years ago that like for me it's more, I can find more competitive and interesting for me stuff than just writing the code. So I have a code writer, a developer who make it for me.
00:10:41
Speaker
Okay, okay, sounds good. Thank you. And if we are moving farther, and my next question is definitely about how you determine the appropriate level of investments in Google Ads, or any other digital advertising channel, giving like the overall marketing goals and budget constraints that the client might have.
00:11:09
Speaker
Yeah, so in this case, I also use my scientific background because when I came to this market, I found out that a lot of agencies use like, you know, a randomly approach, like how much money should we spend on Google Earth? Let me see.
00:11:29
Speaker
Like estimation, doing estimation. Yeah, you should use a lot of money. Okay, let's work together. But I'm using quite another approach. So first of all, I built a math model, mathematical model for the path of client, of typical clients, like what steps should he do? Should he does before he pay money, first time, second and so on. Then I determined the boundary condition.
00:11:59
Speaker
Like what guy do you need?
00:12:01
Speaker
What budget do you have, like boundaries? And after this, I usually find some parameters that are still unknown parameters, like conversion rate for this audience, like cost per conversion, like cost per acquisition, and so on.
Testing and Privacy in Ad Campaigns
00:12:24
Speaker
And now, at the next step, we do an experiment
00:12:31
Speaker
with small budgets just to determine these unknown parameters.
00:12:37
Speaker
in real life instead of just put random numbers. Like I hear an approach when like, what should we do if we don't know the conversion rate on this landing? Just put eight percent. Why eight? Because it's average. You know, it's like what's the average temperature of a patient in hospital? 36 and six.
00:13:04
Speaker
Yeah, because last time we had five and now we hope that this would be bigger. Yeah I prefer just do a small small budget test and we Definitely determine these parameters and at the last step. I usually use the path of continuous improvement when
00:13:28
Speaker
I think widely that only Google ads and try to improve at least by 10% efficiency on each step. And we know if, for example, we have a first step and we can increase by 10% on each step efficiency, we will have 1.5 times more in general income. So 150% in general.
00:13:56
Speaker
Nice, nice. Now like the way of applying scientific methods and research and marketing was also interesting and I was also always excited to hear about that more and more.
00:14:15
Speaker
And how do you ensure that the clients' campaigns are making the most of the latest AI and machine learning technologies? Are you involving the automated bidding, any kind of optimizations, some budgetary allocations between campaigns?
00:14:39
Speaker
So again, I use a well-known approach in math world. It's called continuous test with verification of control group. When you always use at least two campaigns, it's campaigns that works with different settings for each region of product. So for example, this campaign you set up for
00:15:07
Speaker
like audience-based approach and this campaign you divide by products and when you overlaying overlap this data obtaining from these two type of campaigns the results you can understand like a three-dimensional picture how it works in general and usually I try to use
00:15:33
Speaker
as more experiments as I can. Firstly, just before I create a new campaign, I speak with myself, like, okay, what my experience says about how it should work.
00:15:55
Speaker
I don't definitely know how artificial intelligence in Google works. Moreover, it works different for different clients, for different markets, for different months, for different random parameters. Would you say it's evolving throughout the time as well? What?
00:16:18
Speaker
Would you say it is changing throughout the time? Yes, yes, yes, it's very changing. So I just generate some ideas and run as much experiments, as much tests as I can, usually around 50 or even 60 tests at the moment. So it's kind of A-B testing, but like... Yeah, yeah, yeah, yeah, risk control group.
00:16:44
Speaker
Yeah, at the bigger scale. Yeah, that sounds exciting as well. Definitely how you can make any assumptions if you don't have any data. Having more data just allows you to get more optimal results and more successful marketing campaigns at the end.
00:17:05
Speaker
I use a fast scaling approach when I just start from small amount of audience in small region and I expand this approach even for one project I expanded from one region.
00:17:23
Speaker
in four days to all around the world in 16 languages, four days.
AI's Role and Insights in Marketing Strategies
00:17:30
Speaker
Because we obtained the data very fast and we had an ability to scale incredibly fast. I see, I see.
00:17:42
Speaker
Okay, and if we started to talk about audience targeting and with all that concerns about data privacy on the rise and all the regulations that we see as GDPR, CCPA in California and in other states as well, how are you working to balance the need to effectively targeting and personalization with the need to respect users' privacy and data security?
00:18:11
Speaker
I would say now it's one of the biggest headaches for our industry. But in my opinion, it's a headache of ad platforms.
00:18:27
Speaker
So in my case, it's a headache of Google Ads. I just use the best practice of using this platform. So for example, in Google Ads, you can't collect the data that can determine user precisely. You can obtain the data about
00:18:52
Speaker
where he came from, how many times he touched our ads, his behavioral or on our website and so on. But ID identificators are randomized. So we can't predict a specific person. We can only see on general landscape how it works. So in my opinion, like,
00:19:21
Speaker
I definitely can't change the approaches that Google Ads allows us to use with the help of audience and other types of targeting, contextual targeting and so on. So I just use their tools and believe that these tools already respect data of personality.
00:19:51
Speaker
Yeah, makes sense. And I only can confirm that Google is really very deep into taking care about the data privacy and access to the personal identification information. We at Kinesse build some integrations with Google and some other different platforms as well. And sometimes it makes our integrations
00:20:17
Speaker
more complicated than they get used to be because of those regulations. They don't allow you to see the user ID, the PI, they like inventing different mechanisms that would prevent like any kind of person to access that data at all and then getting more and more hard to work with this kind of information and this kind of data.
00:20:44
Speaker
What do you see as the biggest potential risks and rewards of incorporating AI into marketing and advertising? How are you helping your clients to navigate those issues? We've seen multiple examples where
00:21:02
Speaker
Like in the recent years when the AI was doing something crazy while trying to advertise some of the products on the Facebook or Google or on some other platforms as well.
00:21:17
Speaker
So in general, I do not see any risks as long as it is done thoughtfully. So you can't trust artificial intelligence in 100%. I have seen several attempts to simply like turn on automatic strategy, set a budget and go and like, let's add it with how Google ads
00:21:47
Speaker
can do it for us. But as a result, always it's just a waste of money. I usually say to after attempts like this, I usually say like, it would be more efficient to just burn this money, at least we receive warmly.
00:22:08
Speaker
warm and so on. So as for me, artificial intelligence in every platform is like a sniffer dog. As long as you give it the right guidance, it will look for you the best clients in the market. But as soon as you let go of the leash, it will run away with your money. It's like a sniffer dog. You should
00:22:34
Speaker
Right? Like, you know, the approach... Always control, always being in control of what is going on. Yeah. Give the ability to find but control how it works for you. And for the benefits, I think this incredible era, artificial intelligence can present customers who have never even looked for your product as show
00:23:03
Speaker
and show ads to these persons exactly at the right time. I have a very good example from my own advertising. I run ads for my online school.
00:23:19
Speaker
Previously, I used Target for very precise keywords like Google Ads courses, Google Ads Education, how to set up Google Ads. But several years ago, I first tried to give an ability for Google Ads to find me
00:23:44
Speaker
as many potential students as it can. And I found out that Google has found for me
00:23:55
Speaker
I think at this month, 20 times more students and with absolutely unexpected for me search terms, like search term, for example, I very good remember the search term because after the search term, this user bought my course and write to me like Igor.
00:24:18
Speaker
You teach us that we should look precisely at keywords, but I found you with search name, Facebook blocked my business manager. What should I do? And Google showed me like a Google at school figure of whiskey. Why you use like Facebook? It's not your target. And I said to him like, yeah, that's not my keywords, but you are my audience.
00:24:48
Speaker
Yeah, we call it long tail terms, so unexpected. For example, if you need for, I don't know, service of polishing your car, you can previously search for a lot of scratches on the car, what I should do, how to do it by myself. Okay, man, look at this.
00:25:15
Speaker
Well, yeah, that's impressive. And yeah, none of the latest technologies that are being released now on the market, all this AI stuff, Chandra PT and some
00:25:30
Speaker
some other software that allows you to do those crazy stuff that were just unimaginable like even a year or two ago. One guy asked me like Igor how Google knows that I should see ads of this restaurant of this pretty posh restaurant luxury restaurant in the center of Kiev I said
00:25:54
Speaker
like how Google knows that I'm a potential client to this restaurant. So I asked him like, what application are you used for navigating when you drive around the key? Google Maps.
00:26:11
Speaker
That's why, because Google can see what exactly places in the city you visited. Where do you live? Where do you work? That's why.
00:26:26
Speaker
Nice, nice. And if you're talking about those new technologies and new achievements and new products in the market, how do you work to stay up to date with the latest developments in digital advertising and the ad tech in general?
00:26:44
Speaker
So I regularly attend a lot of conferences and the most interesting is that I carefully listen not speech on the stage but communication behind the stage because in my experience you can learn a lot from people
00:27:03
Speaker
months before it gets posted on an industry blogs, on social media, on some YouTube channels and so on. So I'm a really communicative person and I really enjoy to find new people, to speak with them, to share my ideas with them, to carefully listen about their ideas.
The Evolving Landscape of Digital Advertising
00:27:25
Speaker
And I would say that it is
00:27:28
Speaker
If we want to really to be up to date, we should attend as many conferences as we can. Oh, nice. Yeah, this is a good source of information. And can you provide any tips of like, you come to the conference, what do you do? Like, how you talk to all those people?
00:27:47
Speaker
So usually, as for me, the best time to start conversation during the lunch when all of us just eat and you just sit down near some interesting person and like, oh, tasty meat, isn't it?
00:28:09
Speaker
something absolutely unexpected. And after this, I find out a very interesting approach when you use some
00:28:24
Speaker
For example, if it's a conference about advertising, and I usually provide some interesting case, like just for start conversation, like a case from my practice with funny results like humor. Humor is the start of all.
00:28:46
Speaker
nice yeah yeah just some anecdotes related to the topic of the conference or something from your experiences definitely will open up almost every event
00:29:04
Speaker
Yeah, you'll definitely get more out of that conversation and out of that person. Nice, nice. Yeah, those tips on how to expand your network is just invaluable. And we're coming to the end.
00:29:22
Speaker
I have only one last question about the future of the digital advertising and how do you see it and how do you see industry evolution over the next several years?
00:29:37
Speaker
So I would say now we have a time when if or you change or you die. So it's regarded agencies, regarded experts, or regarding product companies. So all the approaches, including approach how to build the structure, for example, of department of demand or of agency, no longer works.
00:30:05
Speaker
So one person with the right mindset now can replace an entire team because
00:30:15
Speaker
because as for me, it's just my opinion. One of the most important thing in advertising now is risk tolerance, the desire and ability to take risks and test come out on top. So for example, my most effective projects took off because of the risk I took
00:30:40
Speaker
doing unusual tests, unexpected tests, like crazy tests. And nobody else can, usually nobody else, especially the average employee of the company can, definitely can't take this risk because it's like, you know.
Google Ads Courses for Entrepreneurs
00:30:58
Speaker
Yeah, because he will have the guide rails, he will have regulations, he will have the approval process and it is a lot of bureaucracy.
00:31:08
Speaker
Usually, these employees never spend their own money on advertising. They exactly don't know how to spend. For example, I have experience when in one day, first January, I wasted near $10,000 several years ago just because I made a mistake in the setup of my campaign, my own campaign.
00:31:34
Speaker
And while the others, as we say, were talking carefully, deliberate steps, I simply overtook them at the turn and speed away with ability to take risks. Nice, nice. Yeah, and by the way, you mentioned several times some courses that you have on the internet. Could you please tell us about them?
00:32:03
Speaker
Yeah, so it's my already fourth edition of my courses in English. Now it's course only in Russian and Ukrainian languages. It's Google Magic 4.0 now and I think in two months, near May,
00:32:24
Speaker
I will create an English version of this course. The main advantage of this course, this course is created by an entrepreneur for entrepreneurs. So it's not for freelancers, not for employees, not for people who...
00:32:46
Speaker
create ads for another ones. It's course for people with entrepreneurial mindset who run their own projects, who run their own companies. So for
00:33:02
Speaker
For persons who, for whom, first of all, their time, because, like my approach, I give only necessarily knowledge, information, the point, the most critical point of growth of your ads.
00:33:25
Speaker
necessary minimum to achieve the maximum success. Yeah, it sounds good. Once you have this, once you have this translated, let us know, we will like post it every way as well. Okay, thank you so much for joining me today for this interesting conversation and have a good rest of your day and see you everybody later. Thank you. I really enjoyed our conversation. See you. Thanks. Bye.