Kraft Baby is a baby products brand that sells strollers, car seats, playpens, high chairs and children's bikes through its own online store, kraftbaby.com. The biggest search in its category, “bebek arabası” (baby stroller), is made about 49,000 times a month in Turkey, and the first page is contested by large marketplaces and the other brands in the category.
When we started in October 2025, rankings were falling: the number of keywords for which the site appeared in Google's top 3 had dropped 29% within that month alone. Seven blog posts on pregnancy and baby health accounted for 19% of estimated organic traffic. In the searches where sales happen, the category pages were behind: position 5 for “bebek arabası”, 12 for “bebek bisikleti” (children's bike), 18 for “park yatak” (playpen).
Our job was to grow sales from organic search, not visits from organic search.
A year later organic sessions were down while organic orders and revenue were up, and Kraft Baby had moved from position 5 to position 1 for “bebek arabası”.
Problem: The Traffic Was There, the Sales Were Not
The site's organic visibility had built up in the wrong place. Its most-visited blog posts answered pregnancy and baby health questions; the people searching for those were not about to buy a stroller or a car seat. The traffic report looked full, and the orders did not follow.
For searches with buying intent the picture was reversed. Category pages sat outside the top 3, and for most keywords several pages competed at once: the main category, a subcategory and a blog post on the same subject got in each other's way in the same search. In this situation, known as keyword cannibalization, Google cannot settle on which page to show and none of them reaches the position it deserves.
In short, we had to solve three problems at the same time:
Stop the decline in rankings
Move category pages into the top 3 for keywords with buying intent
Bring pages competing for the same keyword behind a single target
Strategy: One Page for Every Keyword
Instead of running separate workstreams, we built one map: which page should rank for which keyword. Every piece of work, from titles to internal links and from backlinks to blog topics, followed that map.
1. The map first: the KPI keyword set
In the first month we sorted the target keywords into groups by category and subcategory; each group corresponds to a single page on the site. We reviewed the list with the brand and set aside keywords for products the site does not sell. Our criterion was not search volume but whether a keyword matched a product sold on the site, because the goal was sales, not traffic.
2. The strategic difference: resolving cannibalization with internal links
Across seven category clusters we mapped 58 anchor-to-target pairs and implemented them on the site: which page links to which page, with which keyword. The aim was to show Google a single candidate for each keyword. Subcategory pages kept their place in the map; each became the target page for its own narrow keyword, such as “yenidoğan oto koltuğu” (newborn car seat).
3. Sending backlinks to categories, not the homepage
In the monthly placement programme that began in November 2025, more than 40 articles were published by the end of September 2026. All of them linked to one of nine category pages using keywords from the map; not a single link went to the homepage. The page that needed authority was not the homepage but the category page competing with marketplaces for a keyword like “bebek arabası”.
4. Tying the blog to sales
We did not make it a goal to win back the traffic that was draining from health posts. In new content we gave priority to topics connected to a product, and we linked blog posts to the relevant product and category pages. The blog's job was no longer to collect traffic but to support the category page.
Implementation: From the Map to the Page
Technical foundation
In the first weeks we closed the errors found in the site crawl: URLs returning 404 or still on HTTP, duplicate and multiple H1–H2 headings, missing titles, H1s and image alt text. The crawl is repeated regularly; every item it finds is opened and closed as a separate task.
Category and product pages
In January 2026 we rewrote the category titles around the target keyword in the map. In March we moved the category H1 to the top of the page, and in April we adjusted where the category text sits on the page. In July we renewed the content of the children's bike and car seat categories. On product pages we rewrote the descriptions of priority products, and in October we prepared product-level FAQ sections.
Internal linking
For the stroller, car seat, playpen, play yard, baby bouncer, children's bike and high chair clusters we wrote down three things for every link: the linking page, the anchor text and the target page. For example, the travel system category was linked to the main stroller category with the keyword “çift yönlü bebek arabası” (reversible stroller), and the blog post comparing infant carriers and bouncers was linked to the bouncer category with “ana kucağı bebek”.
Backlink placements
Three to six articles were published each month on news and lifestyle sites. At the start of each month we chose from the map which category page each article would link to and with which two keywords, and submitted the selection to the brand for approval. Most placements went to the car seat, high chair, stroller and children's bike categories.
Blog
The blog kept receiving new content every month, but the choice of topics changed. We picked titles from long-tail questions connected to a product: “From What Age Can a Baby Use an Umbrella Stroller?”, “Car Seats That Pass the ADAC Test”, “Until What Age Is a High Chair Used?”. We reviewed the performance of existing posts one by one and linked the posts to the relevant products.
Timeline
October 2025: Technical crawl and first fixes, KPI keyword set
November 2025: Monthly backlink placements began
January 2026: Category titles, long-tail keyword set and blog titles
February–April 2026: Internal linking decisions, category H1 and text placement, cannibalization analyses for strollers and car seats
July 2026: Category content for children's bikes and car seats
October 2026: Product-level FAQs
Challenges Overcome
Four months without movement
From November 2025 to the end of February 2026 the number of top-3 keywords stayed between 1,017 and 1,151. In the same months organic sessions were well below the previous year; the gap was 40% for November–January. The easiest move would have been to chase traffic back quickly with high-volume informational content. We did not; we stayed with the map and the category pages. Movement began in March, and by early July the number of top-3 keywords had more than doubled from the starting point.
“Bebek arabası” against the marketplaces, with a DR of 42
“Bebek arabası” is the biggest and most crowded search in the category. The first page includes three large marketplaces with a Domain Rating between 86 and 90; when we started, Kraft Baby's DR was 42 and its category page was in position 5. We set out to close that gap on a single page rather than across the whole site. Eight of the placement articles linked directly to this category page. With the 18 internal link pairs in the stroller cluster we separated the keywords the main category should rank for from those that belong to subcategories, and we rewrote the page title around the target keyword. The page reached position 2 in June and position 1 in October.
Results
Rankings stood still for the first four months, and movement began in March 2026. We measured the effect on sales over February–October 2026 against the same months of the previous year, because GA4 e-commerce data before February 2025 was incomplete. Over that period organic sessions fell 10%, organic orders rose 28% and organic revenue rose 86%. The increase was not spread evenly across the months; the strongest period was August–October.
During the same period the brand's advertising investment and its sitewide sales also grew. That is why we use only the orders GA4 attributes to the organic search channel here. Average order value also rose 45%; part of the revenue increase comes from that.
More orders from fewer visits means the organic conversion rate rose 42%. The traffic loss was concentrated in content with no connection to a product: estimated traffic to seven health posts fell 76%. Over the same period combined estimated traffic to the play yard categories rose 85%, and we moved from position 18 to 4 for “park yatak” (playpen).
The number of top-3 keywords more than doubled from its starting point by July 2026, then gave part of that back; on 1 November 2026 it was 82% above the level of a year earlier. The total number of keywords in the top 10 stayed the same. What changed was the share of those keywords in the top 3: from 31% to 57%.
Domain Rating rose from 42 to 50 in a year; over the previous twelve months it had gone from 39 to 42.
For “bebek arabası” the category page moved from position 5 to position 1; in the crawl of 2 November 2026 it ranked above the three large marketplaces on the same page.
Goal / Metric | Result |
|---|---|
Organic revenue (February–October 2026 vs. the same period a year earlier) | Up 86% |
Organic orders | Up 28% |
Organic conversion rate | Up 42% |
Organic sessions | Down 10% |
Top-3 keywords (November 2025 → November 2026) | Up 82% |
Domain Rating | 42 → 50 |
“bebek arabası” (baby stroller) | Position 5 → 1 |
“park yatak” (playpen) | Position 18 → 4 |
“bebek bisikleti” (children's bike) | Position 12 → 4 |
Kraft Baby now has a map that says which page competes for which keyword and that can be applied to every new category.
Takeaways for Brands
Aim for sales, not traffic. Orders can rise while organic sessions fall; what matters is that the searches you are visible for carry buying intent.
Assign one page to every keyword. If a category, a subcategory and a blog post compete for the same keyword, none of them reaches the position it deserves; internal links should say clearly which page is the candidate.
Send backlinks to the page you want to rank. The page that needs authority most is the category page competing for a competitive keyword.
Tie the blog to the product. The criterion for choosing a topic is not search volume but whether there is a product or category the post can link to.
Stay with the plan in the first months. On this account rankings stood still for four months; results came from the fifth month onwards.
Today Kraft Baby is in position 1 for the biggest search in its category and sells more from organic search with fewer visits than a year ago. The next goal is to bring the same map to a result in the high chair and umbrella stroller categories.