In this material, we outline our company’s standard project development cycle.
1. The idea and concept
Everything starts with an idea and a project concept. The idea encompasses what the project will address and the problem it will solve. What audience will it target? The concept explores more specific aspects, such as the project’s functions and roadmap. It’s crucial from the outset to lay the foundation for the project’s development, determine what materials will be included and where, identify information sources, and allocate responsibility.
When forming the project’s idea and concept, we consider monetisation from the start. We typically pre-select affiliate programmes for inclusion, embedding their links from the project’s inception. These affiliate links must be helpful for our users and relevant to their overall content.
2. Information Architecture
The information architecture is based on a semantic core. Usually, we identify several high-frequency queries related to the topic and collect all tail search queries from them.
We cluster the collected search queries, grouping them by meaning, semantics, and user expectations. The objective is to form future clusters, with each cluster containing several search queries the future page will address.
Next, we perform a grouping of clusters, which results in the site’s future categories and subcategories. By collecting all the search queries, we can understand real user queries, cluster them correctly, and then group them.
For example, if our future website focuses on the UK, you might find queries like:
- UK rivers
- UK rivers map
- rivers in the UK
- rivers in UK
- how many rivers are there in the UK
- longest rivers in UK
Such queries would fall under the ‘uk rivers’ cluster – an article about all the rivers in the UK. In the same way, queries responsible for lakes, parks and other natural features can be grouped together as long as they share the same search pattern.
After that, the essential clusters like:
- UK rivers
- lakes in UK
- national parks in UK
Will form the site category – “UK nature”.
In this way, the collected and clustered semantic core influences the category structure of the future site.
In shaping the information architecture, it is crucial to pay attention to several points.
А. Determine the type of content required on the page responding to these queries. It is imperative to look at the search query results and determine what kind of content ranks at the top. For example, if only online shops appear in the search query, the request doesn’t suit us, as the user’s expectations are different. Excluding this kind of query from the information architecture is essential, as our sites are informational and wouldn’t be able to rank at the top in this case. We should always look at the search results and consider the user’s intent with their query.
For example, if the top material is in the form of a rankings page, it would be necessary to generate similar content. If the returned top results required graphics or video content, then similarly our material must also be tailored towards that output.
B. Ensure the query isn’t a navigational query. This is when people ask for something specific about a company or a brand. In most cases, the answer to such a query is immediate in the first place within search results, and the search depth is relatively low. This means that even if we reach the top 5 or even the top 3 for this query, the traffic will generally settle on the first position holder.
C. Pay attention to the number of main pages of sites in the top 10 and the competition for this query. Our SEO team collects key metrics from the top for the queries included in the semantic core.
3. Domain name
A domain name is vital for the development of a project. In most cases, we use drop domains, which are previously registered domains with existing content, external links and ranking.
This can present risks, as previously registered domains could potentially be filtered out, despite high SEO metrics.
However, we adhere to this strategy, and our domain name analysts carefully sift through the thousands of domain names released daily. We evaluate their spamminess, quality of external links, and existing subject matter. We are then prepared to bid at auctions and spend a reasonable sum for valuable domains.
4. Restoring the old structure and link profile
We know that many people who hijack drop domains restore old content to them. We avoid this approach due to potential legal issues. Instead, using various SEO services, we extract the link structure that led to the old site and its menu structure.
We connect the old pages to the new sections and pages of the site via 301 redirects, signalling to search engines that the site’s subject matter has not changed, but the site’s structure has.
This step is also necessary to ensure that old links leading to the site don’t return a 404 error. We restore the old pages and write new content on these pages that are relevant to the queries and links.
5. Website design
At this stage, we already have the site’s information structure, and it’s time to select a suitable website template or create one from scratch. In most cases, we use a combined approach, taking a proven template as a base and customising it to fit the appropriate topics and information structure.
We select the template based on the size and depth of the menu, the type of content we have and how it should be presented. Finally, we create a logo for the site.
6. Terms of reference for texts
After compiling the semantic core and clustering, it’s time to create the terms of reference for the texts. We have to address the following questions:
А. What type of content should be on the page? This information is obtained from the top search results.
B. Determine the future paragraphs of text with h2-h6 tags. This information is also drawn from the top and competitors. If the content already ranks for the keywords we want, we can examine how they are doing it.
C. Determine the optimal length of the text.
D. Identify the questions the text should answer.
E. What other queries for which competitors’ pages are ranking? This allows us to expand the semantics which is already collected and cover a larger number of queries.
7. Writing texts
Once the terms of reference for the texts have been prepared, they are sent to the content department. This step involves working with information sources and creating new, unique content that responds to user demand – a relatively routine process.
8. Photo Editor’s work
Given our numerous projects and the daily publication of hundreds of materials, the work of a photo editor is essential. Their task is to select appropriate images from photo stocks, process them and work with AI image generation services.
9. Data scraping
Many of our projects involve catalogued data or data aggregated from multiple sources, such as posters, event data, exchange rates. Our data scraping department is busy collecting and processing vast amounts of data.
We have our search engine robot to collect data. Programmers then write regular expressions or parse HTML pages for structured data using CSS selectors. Although it’s a routine task, it’s an excellent opportunity for junior developers to learn data parsing in our team.
10. Posting and importing content to the website
Once the first materials for our new Internet project are ready, we import them into the website. Before importing, it is crucial to pre-link the texts and prepare schema diagrams of the materials where possible. Then, we import catalogued data or large arrays directly into the database or via XML-RPC / WP-CLI (WordPress).
11. Link building and analytics
As soon as we publish the initial content, we begin building up the project’s link mass. We analyse the search terms for which the site is starting to rank. If some articles lack link mass, we strengthen them with external and internal links.
Next, the project enters a routine phase of searching for low-competitive queries and situational queries to gather traffic and external links.