HOW TO SET UP GOOGLE ADS, SEO, AND MARKETING CHANNELS FOR AN IT SCHOOL

This is a case study about how we increased the leads by 875% in one year in the field of IT courses in the USA.

Client

Learnix Center is an education company with a government certification in the USA. It focuses on training IT professionals and offers courses in software testing, cybersecurity, and digital marketing.

The company has been operating since 2011 and was founded by professionals with 25 years of experience in this field. Learnix Center not only teaches students but also helps them get a job by supporting resume writing, interview preparation, and job search strategies.

Google Ads

Audit and Strategy

We started working with the client in August 2022. At first, we decided to focus on search advertising in Google Ads.

Before working with us, the client had a $1,700/month budget with previous contractors and was getting an average of 59 conversions per month. With us, over two years, the budget increased to $4,500/month, and the number of conversions grew to 226. By “conversions”, we mean form submissions, phone clicks, calls from ads, and other target actions.

Google Ads Results

> 2.5 times Ad budget increased

3.83 times Conversions increased

by 30% Cost per lead (CPL) dropped

Before we started, we did a full audit of all ad campaigns. We found several serious mistakes made by previous contractors:

• added all possible keywords from different areas, including brand keywords, into one campaign;
• no keyword grouping;
• used only broad match type for keywords;
• no search term cleaning;
• few ad extensions were added;
• no remarketing.

Since the high season was starting, we decided not to fully stop the old campaigns during the first month, because they were already trained and bringing some leads. We only fixed the most critical issues in them and launched new campaigns using our own processes.

In the first month, we set up 5 new search campaigns for the client’s priority courses.:

• Email Marketing
• Software Testing
• Marketing Automation
• QA
• IT Training

We used only hot and specific keywords with words like “training”, “course”, or “classes” to get fast and guaranteed results during the time of growing demand. To filter out irrelevant or low-quality audiences, we carefully researched the keyword meanings and collected negative keywords (like “free”, “diy”, “mba”) before launching each campaign.

We also launched with smaller budgets:

• A search campaign with more general keywords that the client provided;
• A display campaign for people looking for jobs, to get more reach;
• A Performance Max campaign to test Google’s automation and artificial intelligence.

Analytics

At the start of working with clients who already have an active ad account, we first check if Google Analytics 4 is connected and set up correctly. We also check if conversions are tracked properly on different devices and if they are correctly marked as primary or secondary in Google Ads.

The main way users convert on the website is by submitting a form. When we tested this goal, we found it was not working right. This hurt campaign optimization and made the stats less accurate.

We set up a new, correct form submission goal and made it the primary one in the ad account. The old one was changed to secondary. We also set up micro-conversion tracking to better understand user behavior.

We also kept the “Calls from ads” conversion among the main conversions. It tracked all phone calls made through ads using a special Google forwarding number.

Optimization

In each client account, we regularly check performance, analyze how ad campaigns are working, and do basic optimization. This includes:

• adjusting budgets;
• reviewing search terms (removing irrelevant ones and adding useful ones as keywords);
• reviewing Google’s recommendations (applying good ones, rejecting bad ones);
• running cross-negative keywords between campaigns if their search terms overlap. This often happens when campaigns target similar keywords and leads to incorrect stats;
• working with keywords (pausing poor ones, changing bids or match types);
• adjusting bids by location, device, or audience;
• improving ad texts and ad assets.

Testing and Marketing Hypotheses

Besides basic optimization, we spent a lot of time running different tests to reduce the cost per conversion, for example:

• Tested Performance Max campaigns. They brought in cheap leads, but most of them were not valid;
• Display campaigns for broad audiences;
• Smart bidding strategies;
• Different landing pages. For some courses, traffic from the homepage converted better than from custom landing pages;
• Separate display remarketing campaigns for different courses;
• Created campaigns for new keyword groups and new courses added to the site.

The client worked on the website’s CRO. They improved the design, showed why their courses are the best, and added new calls to action. This also helped improve the website’s conversion rate and the performance of ad campaigns.

Thanks to the joint effort, the conversion rate grew from 3% to 12% at the start of 2024.

Challenges We Faced

Irrelevant Search Terms

The main problem we faced during ad campaign optimization was a huge number of junk and irrelevant search terms. For example, searches for free courses, training with future job placement, or courses from competitors.

Even using exact match keyword type and checking the keywords with negative words before launching campaigns didn’t help us get rid of them. Junk terms kept showing up out of nowhere.

So, most of the time during optimization was spent checking every campaign’s search terms carefully. We regularly added negative keywords to remove all junk and updated the global negative keyword list at the account level. This helped stop junk terms from appearing in other campaigns too.

Seasonality and Low-Quality Leads

Some percentage of low-quality leads in this industry in the U.S.A. is normal and something you can work with. But in March 2024, the performance of advertising campaigns dropped sharply—lead quality became much worse, and the total number of leads decreased. Bots started filling out website forms with fake phone numbers, or the leads were underage people (under 18) who couldn’t pay. There were users from third-world countries, even though the campaign was targeted at the U.S.A. Others were just not interested—they were looking for free education or were only curious about the offer.

This happened because of the seasonal slowdown and because the smart bidding strategies couldn’t work well in these conditions anymore. They were no longer able to find high-quality audiences at the set target CPA (cost per acquisition). Instead, they started bringing in questionable and invalid conversions just to stay within the CPA limit, ignoring the quality.

We made several major changes in the account to fix this. These steps helped improve the lead quality and even lowered the cost per lead compared to peak season:

• We added countries with the most bot traffic—India, Bangladesh, and Nigeria—to the list of excluded locations, to make sure Google wouldn’t show ads there.
• We restarted campaigns using manual bidding strategies. We paused the old ones and launched copies to retrain the system from scratch and reset the audience targeting algorithms.
• We set GEO targeting to only 5 USA regions where students can pay in installments or apply for a free course grant through the WIOA program.
• We reviewed all ad texts and extensions to make sure there’s no mention of “free education” and that it’s clear the courses are paid.
• We used the landing pages that had the highest conversion rates during our testing phase.

Paid Social

Where did we start?

We already had advertising on Meta Ads. With us, Reddit Ads and LinkedIn Ads were added.

A full-time marketer was responsible for Meta Ads. Mostly, they boosted posts to a general audience.

We started working with Meta Ads by:

• Auditing event tracking in Meta Event Manager (we found and fixed wrong form tracking events and added clicks on phone and email).
• Updating ad creatives (in the education field, it’s important to tie them to dates).
• Creating and launching a new advertising strategy.

Since this is a complex case with PPC + Paid Social + SEO—and our readers are also interested in challenges—we’ll add a few more details 😉

Form Tracking

The website is built on WordPress, and the forms are added using plugins.

For example, the popular plugin Contact Form 7 works in a special way. To track events correctly, besides using a simple trigger in Google Tag Manager, you also need to “listen” to all AJAX events in the DOM model and specifically catch the wpcf7mailsent event for a successful form submission.

At LearnixTreeCenter, this plugin was used. Additionally, CFDB7 was installed for backup, and all forms were saved in the admin panel.

The Marketing Link team has worked with WordPress since 2016 and knows exactly what can go wrong with tracking events. For example:

• WP Forms (lite)—to track it properly, you need to adjust the plugin settings;
• Elementor Form—it works fine, and the paid version has form backups (for SEO clients, we connect our paid license if they don’t have one);
• Gravity Forms, Ninja Forms—after the latest updates, they are tracked using standard methods in Google Tag Manager.

We immediately updated the work on creatives.

Advertising creatives

In social media, creatives get “burned out”, so we started with the structure, created a shared folder on Google Drive and a Google Spreadsheet, pulled out the most effective ones from the previous period, and planned to add new ones every month.

Paid Social Strategy

Priorities changed during the work process. For example, in summer 2022, the main focus was QA, and also more broad Marketing Automation. By the end of 2024, the focus moved to a new CyberSecurity course.

We changed the structure of the ad campaigns. Instead of boosting existing posts for a general audience, we launched:We changed the structure of the ad campaigns. Instead of boosting existing posts for a general audience, we launched:We changed the structure of the ad campaigns. Instead of boosting existing posts for a general audience, we launched: We changed the structure of the ad campaigns.

Instead of boosting existing posts for a general audience, we launched:

— Instant Form campaigns for specific target audiences;

— Website conversion campaigns for specific target audiences;

— Boosting post campaigns for remarketing and people who interacted before

To reduce random Instant Form submissions (in this ad type, Meta auto-fills the user’s name, phone, and email, so they don’t need to type it from scratch), we added extra questions to the Instant Form—the same ones as in the contact form on the website.

We chose Meta Ads as a secondary source of leads. The main one is Google Ads because the quality of conversion from lead to student agreement is higher with Google.

Because the demand for QA training in IT is not stable, we increased focus on targeted ads during the months with fewer leads.

In the summer of 2022, we created a more detailed remarketing strategy:

• For people who left their contact info (to encourage them to make a purchase).
• For people who visited the site but had only 1-2 sessions (to bring them back).
• For people who visited the site 3+ times or interacted with a form (to spend a bigger budget on a more relevant audience).

At first, we selected interests like:

• Testing
• Software Development
• Programming
• Information Technology

We updated and expanded the targeting during the process. We also split ads by placements:

• Facebook
• Instagram Feed (posts)
• Instagram Stories + Reels

We worked closely with the Sales team at LearnixTreeCenter. They helped us adjust the ads and focus more on Facebook. On Instagram, we only kept remarketing campaigns.

In spring 2024, we decided to add Reddit Ads as part of the “7 touches” rule—mostly for remarketing and some specific subreddits, to bring back students who were unsure.

In fall 2024, we planned LinkedIn Ads for a narrow audience (we can’t talk about the details now) connected to GOV. LinkedIn Ads allow for classic types of campaigns:

• Website traffic
• Lead forms
• Message ads

It also lets you target by job titles, companies, and extra interests.

Optimization

While optimizing social media ads, we also segmented groups by:

• Age (based on CPL—we asked what age converts better to sales);
• Gender (only if the difference was more than 50%);
• Location (Meta doesn’t show lead data in reports if you have several locations in one campaign);
• Placements (we already split Facebook, Instagram, Instagram Stories—and during optimization we also separated Desktop/Mobile or turned off Marketplace, etc.).

The Sales team at LearnixTreeCenter had data not just on leads, but also on which creatives got those leads. For example, we found out that two speakers converted better than the others. So we recorded extra video creatives with them.

Also:

• We optimized bids and budgets based on CPL and CAC;
• We cut spending on campaigns that didn’t convert or were too expensive;
• We reallocated budgets based on campaign performance;
• We planned extra campaigns if they could bring more profit.

Challenges We Faced

Leads and quality

In 2022, we automated all leads from social media, so there was no need to go into the Lead Library in Meta Ads, and the Sales department received the lead not only by email but also in a Google Spreadsheet.

The difficulty was in comparing effectiveness and the lack of a full funnel path. It was hard to compare what was better—100 leads from social media for $15 each, or 20 leads from Google Ads for $75 each.

The Marketing Link team helped with HubSpot consulting (certified specialists and partners), introduced and explained CRM data entry standards, and also helped connect marketing data from the top of the funnel (clicks, leads) with the bottom—sales.

Competitors and Budgets

In Meta Ads, one of the competitors seriously overheated the ad auction with large ad budgets.

Using the Meta Ad Library tools Transparency Center, we looked at their creatives and strategy, and then split our target geos into several dozen ZIP codes, and focused on more targeted ones.

We also noticed that ads in Meta dropped in performance after 2-3 days of running, and further inquiries were much lower quality. We started a strategy of daily ad campaign planning and scheduled stops to reduce spam.

There was also a special discrediting campaign on TikTok related to food testing (Learnix Tree Center trains software testers). We think it was meant to overload Sales managers with fake leads. We helped find a similarity pattern in Google Analytics 4. Even though the spam ads were masked under our Google Ads UTM, we tracked them and made a systematic change to UTM so the manager could know which leads were fake.

And additionally thanks to Ads Transparency Center від Google managed to find a new hypothesis about who was involved. The service shows not only active ads but also the name and country of the advertiser.

It turned out that in the U.S.A. market, through a legal entity in a European country, a company is very actively advertising (800+ active Google ads), and this company had past names related to a billion-dollar parent company.

The information we found helped Learnix Tree Center report to the necessary authorities. It also helped us focus more on closely tracking requests and focusing on real users who are actually interested in the field, not just making spam orders.

Yandex is investing money in an American company through a Swiss company. Politely, this can be called a money transfer. Impolitely—you understand what this means. Since we are a marketing agency, not a political or legal one, we cannot comment on this. What we can do is give advice on how to compete with a company that has a budget hundreds of times bigger than our client’s budget.

SEO

Where did we start?

We started working on SEO in February 2024. 

Since we were already working on Google Ads and Meta Ads, we knew the client’s business well. We understood the features of the keywords and were confident in the accuracy of the analytics tracking. So we could focus right away on:

• technical SEO optimization;
• on-page SEO;
• off-page SEO.

We started with competitor analysis:

• testpro.io
• tripleten.com

We looked at their SEO strategy for buying backlinks.

SEO Optimization

In our work, we focused especially on Local SEO for the following states:
Washington DC (city), Maryland, Virginia, District of Columbia, as the Learnix Tree Center school has special grants through the WIOA program through Virginia Career Works.

After analyzing the growth of competitor backlinks, the size of their blogs, the number of courses, and their Local SEO features, we chose a strategy of 4 blog posts on the website, 4 full PR articles on external platforms, and 8 forum backlinks.

Technical SEO optimization

Before starting work, the site was redesigned. We have been working with the Wordpress platform since 2016.

We checked and fixed problems with:

• website indexing;
• adding to Google Search Console;
• generating a Sitemap.XML file with extra settings;
• no 301 redirect from http to https;
• no H1 meta tags on pages;
• no Description meta tags on pages;
• updated all meta tags to recommended length for best results;
• wrote manual meta tags for the top 20 pages and set up auto-generation for the blog;
• did a CRO (conversion rate optimization) audit of the site;
• fixed broken phone number links;
• fixed placeholder links for social media;
• installed extra SEO plugins: Yoast SEO and one for speed;
• fixed 404 errors with reCaptcha;
• checked 55 outgoing links and kept only technical tools and trusted sources;
• added Elementor;
• created a blog page structure (the default blog theme didn’t have it);

On-page SEO

First, we made small edits to the content on the service pages. Since we are working on paid ads at the same time, we didn’t need to collect extra keywords for the service pages. When choosing keywords, we focused not only on how often they are searched but also on how well they convert from ads.

For the blog, we collected long-tail keyword phrases and topics, and also took popular topics from competitors.

Comparing the sixth month of cooperation with the second, the number of impressions increased by 5.66 times. A more accurate comparison will be possible after one year of work, YoY.

The growth trend makes us and the client happy.

Off-page SEO

Our company has had SEO expertise in the U.S.A. market since 2018. We successfully work with hundreds of local and national businesses in the U.S.A.

In our off-page SEO work, we focused on high-quality local press releases. To support early growth in the first year, we also used forum links to balance the profile.

In 6 months, we published 24 permanent PR articles on external platforms and created 48 forum backlinks.

Our competitors have over 800 domains. Based on our current strategy and plan, we will reach 260+ domains in the next 12 months. At that point, our SEO services will start paying off, and we will switch to more productive SEO packages.

SEO Challenges

WordPress Theme

The website used a custom WordPress theme that did not support a blog.

We involved partner developers from Marketing Link, added Elementor, and set up the blog properly..

Manual Spam from Competitors

Competitors were leaving fake form submissions on the website to overload the client’s managers. CAPTCHA didn’t help because the requests were made by real people from specific locations.

The client uses HubSpot CRM, and the Marketing Link team has HubSpot Inbound Certification. This helped identify the attackers and move the fake leads to a separate list (basic spam from the same device, same geo, and same IP, just different browsers). This helped reduce the load on managers at Learnix Tree Center.

Results

Overall, when the marketing budget increased and services expanded, comparing the last 12 months to the previous period:

• The monthly marketing budget grew by 300%;
• The number of monthly inquiries increased by 875%.

Conclusion

A complete approach gives results when building a marketing strategy.

At Marketing Link with Learnix Tree Center, we were able to set up a long-term partnership approach. It included not only ad and bid optimization, but also regular feedback and new creatives.

The SEO service helped improve the content on service pages. As a result, the Quality Score in Google Ads got better. That lowered the cost per click and the cost per lead.

We also helped choose the right certification for launching the new CyberSecurity course in the U.S.A. market using a Go-To-Market strategy.

Thank you for reading. Leave a request and we’ll find a way to improve your business.

Order a personal consultation for your education business.