SEO Magic: How Housing grew more than 5x and overtook Magicbricks and 99Acres in 18 months
Ravi Bhushan is the founder of early-stage Ed-tech company, Brightchamps. Before venturing on his own, he was the Chief Product and Technology Officer at the PropTiger Group which runs three real-estate properties: Housing.com, Makaan, and PropTiger. Ravi also held the business responsibility of Makaan.
For the sake of context, Magicbricks and 99Acres, are the incumbents of the Indian real estate business. At the start of 2017, Magicbricks and 99Acres were roughly five to seven times larger than the combined traffic of Housing, Makaan and PropTiger. But, in less than two years, by Sept 2018, the PropTiger Group (for the sake of simplicity, referred to as Housing from now onwards — since Housing is the consumer-facing brand and the largest destination within the PropTiger Group) overtook both Magicbricks and 99Acres and became the largest real-estate destination (in terms of the combined traffic). More importantly, most of this traffic was organic traffic.
How did Housing achieve this? What SEO magic did they perform? Ravi shared his secrets during our conversation.
Ravi’s SEO journey into the SEO world started due to “a traffic accident involving Panda in India”! Housing’s traffic dropped by 70% overnight due to Panda penalty imposed by Google. This put paid to the product and engineering investments that the Housing had made over the previous several quarters. Ravi said, “it is painful to have created a beautiful painting and then realizing that someone has put the painting in the dark room and locked the door”! It is important to not only build good products but also to have a clear-cut strategy in terms of how the traffic, especially organic traffic, will come.
What is the most important SEO activity?
The most important and the most impactful SEO activity is “site structuring”. Site structuring ensures that the website is structured in such a way that Google bot crawling the website is provided with the right syntactic and semantic signals about the importance of different sections, pages and content of the website. Site structuring, therefore, focuses on getting the basics right.
First, company should come up with the right information architecture based on the semantic information that the company wants to communicate. The site should be structured such that not only a human being but Google Bot can also understand the importance of different topics / concepts. Each section (corresponding to a topic / concept) should be provided with the right weightage.
To do this, each section of the site should provide coherent meaning and different sections should not overlap too much with each other. For example, a car website can have hub-pages corresponding to the brands (such as Maruti, Tesla, Mercedes, etc.) and, under those pages, hub-pages for various models of the brand. In addition, car website can have a different set of hub-pages for different categories such as sports cars or electric / diesel cars. It is important to structure these pages carefully (as well as link them to the right product detail pages) so that users as well as Google bit can understand the semantic meaning of different sections.
Second, in order to give the right signals to Google and other search engines, it is important to link the pages in a thoughtful manner. Linking pages indiscriminately and ending up with a spaghetti of interlinked pages (including, for example, linking “Contact us” or “Disclaimer” pages from all pages — a common mistake) confuses Google bot and often leads to pages not getting indexed and ranked properly.
Ravi pointed out that linking two pages is nothing but a vote of trust that one passes from one page to another. It is important to link pages by double-checking them from the business priority point-of-view, keyword point of view (importance of specific topics that the company wants to cater to), and from traffic point of view (what opportunities areas are there and how does the company serve those opportunities).
Ravi pointed out that such site structuring experiments can have short-term negative impact (since Google has to decipher new structure and change the earlier set of indexed pages). For example, Housing experienced 30% drop in traffic in the first month after releasing website with proper site structure. However, the traffic increased by 350% in the second month — so, one can get really good results just by doing site structuring properly.
How can one prioritize across various SEO experiments?
After taking care of hygiene-level activities such as site structuring, it is good to use a prioritization framework that takes ROI into account. One way to compare various experiments is to measure them along four parameters:
- [Volume] Volume of opportunity: quantum of traffic available
- [Quality] Quality of opportunity: relevance of traffic to the business
- [Difficulty] Keyword difficulty: how much is the competition for specific keywords
- [Cost] Cost of experiment: effort required to run the experiment
First are foremost, one needs to be aware of the given keyword universe and the size of the opportunity — number of search traffic volume one can go after. Next, one needs to consider users’ intent — which helps one to understand the ability to engage with and convert the visitors into customers. Third, one needs to factor in the competitiveness for specific keywords — what’s the domain authority of the competitors, what keywords have they captured deeply, etc. And finally, the cost of the experiments corresponds to the effort (engineering, content, etc.) required to do the tasks.
Given the above, the experiments can be ranked using the following ROI metric:
(Volume * Quality) / (Difficulty * Cost)
This can be applied to various types of experiments — related to a set of keywords, adding platform features (such as images or videos on product pages; or starting initiative to collect UGC content, etc.), or even re-building the platform from scratch.
Unfortunately, almost 70% of time SEO activities are related to minor tweaks such as title changes, meta-tag changes, writing more content, etc. However, by taking the ROI-based approach, one can avoid dealing only with tactical stuff and start picking up more strategic initiatives. Of course, the ROI metric is just an indicator; the exact decision should be made by considering other qualitative aspects such as the business vertical, type of website (transactional or informational, e.g.), platform capabilities, etc. Together with the ROI metric, one can arrive at the most effective SEO roadmap.
Using the above framework, Housing was able to pick the right experiments that helped them surpass Magicbricks and 99Acres, both of which had much higher domain authority. Instead of going after every keyword related to Indian real estate and competing with the incumbents on the “head” keywords (that generate a lot of traffic; these were “property in Mumbai”, “property in Gurgaon”, etc.), Housing focused on the housing project related keywords. In addition, Housing realized that a lot of traffic (as well as higher intent traffic) was in the long-tail keywords such as “2BHK property in Goregaon”, “3BHK Sohna Road”, etc.
These decisions (to focus on specific set of keywords, e.g.) have an impact on the site structure and the information architecture. They also have implications on what kind of platform one needs to build — for example, how should search and navigation function, how should mapping and other infrastructure work, etc. These need to be factored in while calculating the cost of the experiment.
Ravi emphasized that the same analysis can be done by early-stage startups as well. For this, he gave an example of PDFdoctor.com, a passion project that Ravi started after leaving the PropTiger Group. PDFdoctor provides tools to work with PDF documents (like merge or split PDF documents, etc.). As one can imagine, there is a lot of traffic for online tools related to the PDF; also, this is a super-competitive space with several companies working on building tools (and doing SEO for them) for over a decade. However, by prioritizing various activities using the above-mentioned framework, PDFdoctor identified the best candidates to focus on and, as a result, started ranking amongst the top three results globally within 6–8 weeks! Therefore, by understanding what the competitors were focused on and what they were not, it is possible for any startup to compete globally (across multiple languages and countries) and come out as the winner.
SEO Building Blocks
Most early-stage founders know that SEO is important but a lot of them are not able to leverage it properly. What can startups do to benefit from the SEO magic?
Ravi suggested that startups have to create the right culture and the right infrastructure in the company to ensure that SEO gets the focus it deserves.
Culture requires understanding the importance of SEO and talking about in senior management meetings and in the company townhall discussions. It also requires empowering a leader to focus on SEO growth and creating a dedicated team or squad focused on SEO. The SEO team should have clear targets and goals (along with dashboards that provide measurability) and the team should be rewarded for both mini-successes and for achieving major milestones. A lot of companies miss this and don’t have anyone responsible for day-to-day handling and growth of organic traffic.
Infrastructure corresponds to infrastructure around log management, A/B testing, etc. This is because SEO should be a proactive game; it should not be played in a reactive manner (reacting to competitors or Google penalty).
Ravi pointed out that Housing had created a separate micro-service for SEO to take care of all technical SEO requirements such as title / header / meta handling, page redirection, content interlinking, content spreading, etc. In other words, all major aspects of technical SEO were supported by a micro-service.
The right infrastructure empowered the developers and PMs to do any experiment they wanted to do in isolation, without impacting the main product roadmap. This enabled the team to run their experiments very quickly and, therefore, become more effective.
The SEO team itself can be part of the product, tech or marketing team; the important part is that (a) it should have SEO-focused developers and product managers and (b) the company should work to ensure that the team doesn’t face roadblocks while taking care of the SEO tasks.
Difference between paid marketing and SEO mindset
Paid marketing requires understanding of various platforms (such as Google, Facebook, YouTube, Instagram, Linkedin, etc.) and the related marketing tools as well as planning (for example, budget planning, allocating spends across different platforms, etc.). Also, one can get instant results (within a few minutes or hours) regarding paid marketing experiments. Unlike paid marketing, SEO requires longer time period to show results and, therefore, more patience. More importantly, organic marketing and SEO requires a different mindset. This mindset demands product DNA and product-first thinking.
Product DNA is important because, ultimately, Google rewards those websites that are loved by users. So product efficacy is the core of organic marketing and SEO. In other words, as far as the organic traffic growth is concerned, a good product wins in the long term.
To imbibe product DNA, one needs to adopt user-centric mindset. As a result, it is important for the product managers (those who are responsible for the overall product roadmap) to start thinking about SEO items as well.
Treat Google Bot as a User Persona
To avoid the conflicts and to ensure that the SEO does not get Cinderella treatment within the organization, Ravi said that we must treat Google Bot as a user persona!
Why is this important? Ravi said, “imagine who would be the most frequent visitor of your just-recently-launched site? It is going to be Google Bot!” Therefore, one needs to take care of the Google Bot by presenting it with the right site structure, making the site easy to crawl and index, ensuring fast response time with high download rates, etc. If Google Bot finds the site friendly, it would rank the site higher and, therefore, tell the whole world about it and generate a lot of referral traffic!
At the later stages, when the site starts generating good traffic (direct and otherwise) and the product / platform has evolved, then the Google Bot user persona should be considered along with other user personas. At this stage, high quality product (that serves the end-users’ needs) is the most important. If the product quality is high and the end users like the product, Google Bot persona also ends up “liking” the product — even if you do a few mistakes. In other words, one can naturally start treating Google as a channel at this stage and focus on prioritizing user-specific activities (instead of resorting to SEO hacks to rank higher in Google results).
Treating Google Bot as a user persona can ensure that the team views SEO empathetically and applies user-centric mindset to SEO activities as well. It becomes easier to align different features and to consider SEO activities as part of the product and tech roadmaps. As a result, product/tech roadmap conflicts get sorted out.
This, Ravi hoped, would be one of the key points takes away from the conversation.
