Hi, I’m Alex.

This page will give you an idea of my professional skills and experience. Enjoy!

 

Career Overview

My career can be divided into two main periods:

Freelance (2014-Present):
Since the end of 2014, I’ve been a top-notch freelancer. I usually work as a remote team member, responsible for paid user acquisition. I help online businesses move their profits to the next level.

Agency Experience (2010-2014):
I worked at one of the top SEM agencies in the country. I started as a web analyst for 8 months before being promoted to PPC team lead for nearly 4 years. With my team, I turned the PPC department from a negative profit to a consistent 5-figure monthly profit in USD.

 

FAQ

What budgets do you work with?

• The budget per client ranges from $10,000/month to $200,000/month.

How many businesses have you worked with?

• As a freelancer: Around 15 since 2014, managing 1-3 businesses at a time.
• As a PPC team lead: Over 700 businesses in 4 years, handling 50-80 businesses at a time.

What channels do you work with?

• My favorites are Google Ads (Search, GDN, Discovery, YouTube) and Meta Ads.
• I work with any channel I believe could be profitable for a client.

Additional Skills:

• I build chatbots for WhatsApp, Messenger, Telegram, etc.
• I set up software integrations, such as passing data from your CRM or chatbot to Google Analytics/Google Ads/Meta Ads, or from an ad system to your CRM via Zapier or by instructing your programmers.

Do you work alone or with a team?

• It depends on the project.

Please, show me some of your work.

• Here are details of the most interesting challenges I solved:

1) Bookimed.com – a globally recognized medical tourism agency

Main goal: generate more profit

Markets: Europe, Asia, US

Challenge #1: increasing the number of leads doesn’t always correlate to an increase in sales

As many businesses do, they tracked the number of leads and CPL (cost-per-lead) as KPIs. Though, when they tried to scale by getting more leads at a target CPL, it led to unpredictable inconsistent results because not all leads are created equal.

Solution:

We moved to SQL (Sales Qualified Lead) and cost-per-SQL as the main KPIs for ad campaigns. To start tracking SQLs, with the help of the client’s programmers, we deployed a system where data from CRM was exported to Google Ads automatically. SQL has a strong correlation with the purchase intent, so getting more SQLs means more sales. It opened a lot of new opportunities to scale and leverage auto-optimization of ad campaigns, including the following:

• Getting more signals of higher-quality leads. We started sending other CRM statuses too. So, for instance, if a user completed more stages on their way to purchase, Google Ads has more signals about this user and tries to get more users like this.
• Getting fast feedback about lead quality allows to
  • leverage auto-optimization algorithms that aren’t efficient for longer decision-making cycles, and
  • run A/B tests for landing pages, ads, and user acquisition funnels that require fast feedback (conversions data) to allocate traffic dynamically or to get actionable data within a few weeks, not months or years.

Challenge #2: long purchase cycle

It may take a few months from initiation of research by a user to getting remuneration for this patient from a clinic -> hard to run A/B tests, and make decisions.

Solution:

Introducing the aforementioned SQL (Sales Qualified Lead) and cost-per-SQL as the main KPIs for ad campaigns solved the issue. We’ve identified the status in the user’s funnel that happens early enough to be applicable for decision-making and highly correlated with the purchase.

Challenge #3: multichannel attribution

When people are in-market for these kinds of services, they interact with the business multiple times. They start from one channel/ad, then use organic search, then return via a video on YouTube. How to attribute just the right amount of credit to a channel/campaign/ad/keyword?

Solution:

There are complicated and expensive tools for this purpose. But I prefer integration with the CRM, and offline conversion tracking with custom attribution. It allows passing this data to the ad account so the algorithm can leverage it and manage bids based on it automatically. And, it’s free.

Challenge #4: get more profit without decreasing current ROI

Even search traffic has a low CR and high CPC in this industry. How to scale when you can’t afford a higher CPC? One of the ways is to increase CR.

Solution:

They used to go for that old approach: sending a user to a page with a catalog of products (clinics, in this case) that could solve the user’s problem. But there is an approach that drives much higher ROI: choose the best solution for a user and present it to them in the most appealing way. See how we did this in the case study.

Challenge #5: misleading reports about revenue (imported via measurement protocol) vs ad cost in Google Analytics

Revenue data in Google Analytics, imported via measurement protocol, is misleading. It’s always shifted because the purchase event is attributed to the date of making a purchase, not the date of click. And the efforts you’ve made and the cost you spent to get this purchase happened much earlier. So, when you analyze the first part – you always get poorer performance than it is. And, when you analyze more recent data, the performance always seems much better than it actually is. Even if you stop the campaign, you may see it “performing well” and generating pure profit for some period after it has been stopped.

Solution:

We started using reports in Google Ads – they attribute the purchase to the date of click, not the date of purchase, so they show you a useful picture. We integrated CRM so it sends data into Google Ads about actual sales and revenue.

 

2) NDA: medical billing and coding courses

Main goal: generate more profit

Market: US

Challenge: unprofitable CPL and low conversion rate from lead to purchase

Solution (in 2019):

Revenue per lead was so small and competition so high that it wasn’t possible to get a consistent profit from ads. So, instead of focusing on decreasing CPL, I focused on increasing the conversion rate.

They used email for lead nurturing. I’ve set up a chatbot for FB Messenger and switched the follow-up sequence to it. It was a “new thing” back then, the delivery rate was close to 100% and the open rate was ~60-80%. It increased the revenue per lead by ~500% -> allowed us to scale from there. Chatbots are more efficient than email for nurturing leads nowadays too.

 

3) NDA: Clean-up software for Mac OS

Markets: worldwide, with a primary focus on the US, UK, CA, AU.

Challenge #1: strong competitors with huge budgets -> high cost of traffic -> thin margins. A need for generating a big number of sales while adhering to strict ROI targets. Consistently and at scale.

Solutions:

1) Systematic testing of a big number of variables at scale

Testing a big number of variables, such as traffic sources, sales funnels, ad placements, ads, and landing pages requires a system.  It eliminates the influence of a human factor on decision-making. Plus, it allows you to leverage AI for an auto-optimization in  Google Ads, Microsoft Ads, and Facebook ads.

How the system looks like:

• set up accurate tracking. Installs and purchases are hard to track accurately via web methods, such as JS and cookies. We configured the backend so it sends the conversion data into ad accounts.
• define KPIs, such as ROI, cost per order (CPO) and cost per install, and,
  • the optimal target value for each KPI per channel (a high ROI doesn’t necessarily lead to a higher overall profit. It often means lower bids -> fewer sales. Optimal targets – ones that lead to the max profit),
  • threshold values and algorithms for making decisions (example: if the amount spent > X and the number of conversions < Y -> pause a channel/targeting option/ad/landing page) based on statistically significant data (see more in this article).
• where possible, I automated the processes of identifying and applying winner variants
• where automation isn’t possible – standardized the processes of making decisions

Then, the only thing left is to analyze what works and what doesn’t and conduct new experiments based on the findings. Like a conveyor.

2) Leveraging tens of traffic sources

I tested any traffic source that could send us Mac OS users. Some of them were obvious, such as Google Search and Facebook on desktop. Others were not that obvious. For instance:

• Mobile search. We just had to tweak the sales funnel: instead of the “Download” button, it says “send the download link to your email”. No competitors were there -> clicks cost almost nothing -> high ROI.
• Bing search. It’s owned by Microsoft – how could it send Mac traffic? Though, I convinced the client to test it out. It turned out that it does have a considerable portion of Mac users – ones that use the Firefox browser, where Bing was the default search engine. Plus, the cost per click is much lower there, so the overall profit was comparable to profit from Google Search.
• Software aggregators, such as Softonic, Cnet
• Pop ads
• Direct deals with websites with high traffic, such as speedtest.net

3) Tailoring the app version to traffic source or landing page/ad’s content.

During one of the tests, we figured out that we could keep the message consistent not only through the ad -> landing page’s content but also during the in-app journey, including pricing options. For instance:

• if the user came from a placement like speedtest.net (the ad is about “how to increase your Internet speed”), the app would focus on features that help to increase the speed of the user’s internet.
• And if they came from a landing page that is focused on “cleaning your Mac to keep your browsing private”, the app would focus on this in the first place.

4) Developing in-house tools to make decision-making easier and faster

The tools show instantly, what a human has to do. For instance: pause this ad and increase bids on these placements by this amount. More examples:

• IPs, placements for blacklists
• pulling data from ad systems into the backend via API
• suggesting winners for A/B tests

Challenge #2: bot installs, bot traffic -> ad systems’ AI receives misleading signals about campaigns’ impact.

Solution:

Offline conversions import for Google Ads, Microsoft (Bing) Ads, and Facebook Ads. With the help of the client’s programmers, I’ve set up a system where info from the backend is sent to ad accounts only when real people installed the app. It gave the right signals to the systems for an auto-optimization.

For other traffic sources, we’ve set up automatic generation of a black list for placements and IPs.

4) Filter.ua – a national leader in water purification systems (retail)

Market: UA

Challenges: 

• thin margin, especially in products of popular brands
• low sales volume of products with a good margin (non-popular brands and private labels)
• hard to scale: volume of sales is restricted by search demand; social and display ads aren’t profitable
• a big number of competitors -> high cost of traffic
• 85% of sales happen over the phone
  • high load on call center (additional expenses)
  • harder to track/attribute multiple touchpoints via different contact methods, campaigns, channels, devices, keywords, ads
  • a high portion of leads are in their initial research stage -> lower CR from call to purchase and sales managers are less satisfied with their job

Solutions:

Of course, I increased the profitability of their search and shopping ads by optimization. It was the low-hanging fruit.

Then, we improved conversion tracking – we moved to offline conversion tracking. Now, we see all purchases and profit (not just revenue) in ad accounts, even those that were made via phone calls or chats in messengers.

But then we started thinking bigger. Why do 85% of users need to talk to a sales manager before they buy? What if it’s because they don’t have the competence to judge: which of the products is optimal for their situation.

Step 1: let’s stop making prospects think

Instead of sending a user to a page with a list of products, we started sending them to a basic quiz. They tell us what they do know, such as how many water consumers are there in their household or do they have limescale on the sink. And then, a sales manager recommends the optimal product for their needs.

Result: much more leads, much lower CPL, a shorter time frame from a click to purchase, and happier sales managers. See details here.

Step 2: what if we replace the sales manager with a robot?

I’ve built a funnel via a chatbot.

The bot asks the same questions and replies with the same answers a real sales manager would. Then, it recommends the optimal product for the prospect. If the prospect doesn’t buy, it sends follow-up messages, and nurtures leads automatically.

Result: it’s the game-changer! This bot is like a sales-generating machine now. We feed it with traffic and it sells products that used to require a lot of research and consultations.

Step 3: let’s increase the profit margin and the number of sales

A potential buyer needs clean water. There is a huge difference in profit margin across brands and types of products that can solve this need. You can sell them a product of a well-known brand for $200 and earn $20. Or, sell a product of a different brand for the same $200 but earn $60. Or, sell them a product of your private label for $180 and earn $90. We’ve focused on the latter.

• When people are on Google, it’s too late for high-margin deals. Most of the prospects are searching for a specific popular brand or model and comparing sellers and prices. Products of our private label are not on their list. At this stage, it’s hard to convince a buyer to buy a different product – the one that would solve a buyer’s needs and have a high margin. We launched ads on Facebook, Instagram, Google Discovery campaigns, YouTube, etc. to acquire the prospects before they go to Google.
• And, there is the second variable: the cost of traffic. Search and shopping campaigns are easy to set up – too many sellers do it. It leads to very expensive CPC -> lower ROI. On the flip side, making a profit on Facebook, Instagram, Google Discovery campaigns, YouTube, etc. for such products is a much more complicated process -> leads are much cheaper there. The traditional approach, such as showing a product and its description in the ad doesn’t work there. You need to warm the prospect up, nurture the lead, and only then do you present your product. Our chatbot does this job perfectly.

Results:

• many more sales because we can generate leads profitably on more channels than our competitors.
• much higher profit on each sale because we acquire prospects before they research other options and sellers. It allows to offer them an optimal solution for both: seller and buyer. The focus is on private label products.

 

5) Petcube – The Smartest Pet Cameras For Cats & Dogs

Markets: US, UK, AU

Channels: Amazon Ads, Facebook (Meta) Ads, Google Ads (Search, Display, YouTube), Bing (Microsoft) Ads.

Challenge #1: generate $200,000+ in 45 days for the campaign on Kickstarter (while adhering to a strict target CPA threshold).

We reached the target of $200,000 in less than a month and generated a magic number of $319,193 before the campaign ended.

Most sales were generated via Facebook Ads. Advertising campaigns were deployed in 3 stages:

1. generating leads 2 weeks before the launch of the campaign on Kickstarter
2. nurturing leads, then sending them the early-bird offers when the campaign on Kickstarter has started
3. running direct response ads a few days after the launch of the campaign on Kickstarter

Challenge #2: Track the impact of online ads on offline purchases in major retail chains in the US

A big portion of purchases happens offline in major retail chains in the US. They share no info about the buyer.

Solution

Buyers of Petcube need to install their app to start using the device. After the installation, the buyer’s info such as advertising ID, phone number, and email becomes available in the app’s backend. With the help of the client’s programmers, we retrieved this info and passed it, along with the purchase value, to Facebook (a.k.a. offline conversions tracking).

Challenge #3: cross-market the box office hit The Secret Life of Pets with the Petcube camera

To mark the launch of Petcube’s flagship product, the Petcube Camera, the client secured a deal with Universal Pictures to cross-market the box office hit The Secret Life of Pets with the Petcube camera. With creative which integrated footage from the film as well as maintaining product focus, the client’s objective was to reach a tech-loving audience in a number of international markets where the product is launching.

I launched ads on Facebook and Google to reach a broad, entertainment-loving audience who are tech-savvy and pet lovers, in the US, UK, Germany, Japan, and Australia.

6) Go IT – web dev, web design, and programming courses

Market: Europe

Channels: Facebook (Meta) Ads and Google Ads (GDN, Discovery, and YouTube).

Challenge: Profitable high-volume (10,000+ leads per month) lead generation.

 

Result:

More than 15,000 leads/month.

Example of one month’s stats in Facebook Ads solely:

 

SEO Studio – one of the top SEM agencies in the country

Challenge: to turn the profitability of the PPC department from negative into a consistent profit

By the August of 2010, the PPC department in my company was generating financial losses. I have been promoted to PPC team lead with one goal: to make it profitable. With the help of the team, I turned the profitability of the PPC department from financial losses into breaking even in a matter of 2 months. Within the following few months, it started generating a 5-figure monthly profit consistently.

Our agency in 2013

5 things I did to achieve this goal:

1) I defined the most profitable segment of our current and potential clients – retailers.

2) Then, we developed our competitive advantages, such as:

Search ads automation for retailers and shopping comparison engines via Google API

It was 2010 and Google Shopping wasn’t available to the local market up until late 2013. So, we developed our own software using Google API and stores’ XML to automate the following:

1. 1. • keyword generation based on a product’s name and category,
      • grouping keywords into ad groups,
      • processing of search queries to find potential negative keywords,
      • ad creation, editing, pausing/activating depending on a product price and availability status changes.

It allowed us to acquire the biggest retailers and shopping comparison engines in UA such as comfy.ua, allo.ua, and ek.ua with $6-figure/month budgets plus more than a hundred smaller businesses.

I delegated ownership of this product to one of my team members in a year.

In-house call tracking software

It was 2011 and there were no call tracking solutions in the local market. So, we developed our own: it replaced the phone number depending on the cookie content.

Real orders tracking

Some of our clients experienced issues such as orders on their website without the intention to buy. For instance, a competitor of the jewelry store made “orders” of many expensive products on a regular basis. Our client’s adult toys shop had orders, made by prankers, especially on a Fool’s Day. Traditional web tracking of orders wasn’t helpful – too much garbage in statistics.

Our first solution (in 2011) was creating an in-house software where we

1. 1. • pulled real orders’ data from the clients’ CRM via XML
      • pulled traffic data (source, medium, campaign, keyword, ad, etc.) from Google Analytics via API
      • matched by order ID, laid out in a form of a table with filters.

Then, offline conversion tracking in Google Ads and Facebook Ads, and measurement protocol in Google Analytics became available. We started using these solutions instead and they are relevant nowadays too.

Custom automation solutions for clients

For instance, traffic and conversion rate for pizza delivery increase a lot during rainy or cloudy weather. We had to adjust bids and budgets according to the weather. It’s not feasible for a human and can’t be factored in by performance-based bidding strategies. So, we took weather.com’s data via their API, processed it on our site, then sent corresponding adjustments to ad accounts for our pizza delivery client.

Store’s Profitability Optimization (Conversion Rate Optimization + UI, UX improvements + taking margin into account)

PPC optimization and automation are only a part of the process. Profitability also depends on how well the website converts, how big of a margin the first few products on a landing page have, etc.

Conversion rate is one of the variables. We optimized it via UI/UX improvements, conducting A/B tests on a regular basis.

There is another variable many retailers neglected: the way they sort products by default. The most profitable approach is to show the products with the highest RPM (revenue per 1,000 impressions) first. RPM = estimated conversion rate * profit margin. So, our clients adjusted their backend to increase the RPM. It increased the profitability of all the traffic we sent to their website.

3) Workflow standardization and optimization

A growing number of clients causes a need for new specialists which causes new expenses. Instead of growing the team, we focused on standardization, automation, and optimization of our workflow.

We introduced SOPs (standard operating procedures) for the most time-consuming and high-cost-of-error processes. The most experienced team members were keeping the manuals up-to-date.

4) Cost-efficient increase in staff resources

Soon, we decided to hire more people to keep up with the growing number of clients. Here are the changes I’ve made in the hiring process:

• Hiring proactive people who are fine with going the extra mile on their own initiative. To filter out such people, I introduced a test potential hires had to pass before they can send an application. It reduced the number of applications from those who saw the position description from 20% to 2%. But from that 2%, each third was great enough to be offered a position. It dramatically increased the quality of resumes I received and saved time on the hiring process.
• I introduced a bonus system. On one hand, it increased motivation. On the other hand, it spread financial risks because part of the salary became dependent on how much a team member earns for the company.
• Involving the team members in the hiring process. Each new hire was assigned a probation period. After the probation period has elapsed for a person, the whole team decided if we want this person to stay with us.
• In 2013, I introduced the second layer in management. The most efficient team members were promoted to team leaders.

 

5) I improved communication with Google reps 🙂

It allowed getting early access to new features in Google accounts that were available to a limited number of advertisers.

Me, playing with the googler in our office

As a result of the implementation of the abovementioned measures, the profitability of the PPC department increased to a consistent $5-figure monthly profit because:

• our agency acquired many new businesses. Clients became more satisfied with our services. It increased the average time they stayed with us, the budget they spent, and the remuneration they paid to us.
• Team members spent less time per task/per client. It decreased the cost of production.