Case study on launching advertising for metal products
Launching search ads and testing display ads
Background
The client sells various metal products: rolled pipes, bronze, aluminum, their alloys, and finished parts—bushings, pipes, and much more. Essentially, it is a metal warehouse with a wide range of products.
During the course of our work, the client had several websites. Later, he launched two more, and as a result, we tested advertising on four websites.
Ad campaigns start
We mainly launched search advertising. We tested banner campaigns minimally, but quickly abandoned them: there were almost no applications, and those that were received were unresponsive and without real demand. Search advertising proved to be sufficient to meet our lead generation needs.
First, we asked the client for a full price list, which helps to see all the products and categories, since the website usually only shows part of the range. Based on the nomenclature, we grouped the products (by item number or name) and compiled keywords.
Next, we worked through each group using Google's keyword planner, added relevant queries, and prepared a campaign. At the start, we always use exact keyword matching to make it easier to work with search query reports. After cleaning up irrelevant queries, we move on to optimization.
There were enough queries in this niche, so we continued to work with exact matches, but then switched to smart bidding strategies. The campaigns showed good results: leads were stable, and the client was satisfied.
Scaling
At the start, the client had two websites, and we created a separate advertising account for each one. Initially, the campaigns were based on the keyword “metal warehouse.” When the question of scaling arose, we smoothly transitioned to adding groups with broad matching directly to the existing campaigns (using a “maximum conversions” strategy). This resulted in an increase in leads of approximately 30% without changing the cost per lead.
Semantic<>core
Smart
bidding
Media
ads tests
Support
service
However, tests showed that the old sites yielded better conversion results. The two new sites were tested for several months, but in the end, the budgets were returned to the old ones. There is a hypothesis that new sites perform worse in Google advertising due to a lack of “trust” and accumulated data.
Support
Спустя несколько лет работы проект перешёл в стадию поддержки: все кампании запущены, они стабильно работают. Наша задача — раз в неделю проверять рекламные аккаунты, следить за техническим состоянием сайтов и кампаний, чтобы всё оставалось в порядке.
Такая услуга в агентстве тоже есть: поддержка уже работающих кампаний без необходимости ежедневного вовлечения.
