Complex work on pik.ru website

Start

At first the customer had a group of websites where pik.ru was only a spreading page. The catalogue of objects was located on pik-estate.ru, whereas the main pik.ru domen was three times more authoritative and cited. This general marketing mistakes made the website lose a variety of advantages, and the company got the insufficient search traffic.

Work performed

First of all, the customer was offered to change a marketing strategy: the website pik-estate.ru was completely transmitted to pik.ru, and the information on the company was referred to a separate corporate portal pik-group.ru.

The main domen content of the company before and after the information transfer:

The ordering of domen structures produced a number of advantages:

• It is a simple and easy to remember address which is further used both online and offline. Much higher associative domen connection with PIK Group, greater authority and customers’ trust (and search engines), higher CTR (click-through-rate) in search results promoted the rise of positions.
• The site ergonomics was also improved by minimizing the number of customers’ clicks up to the necessary information (offers of real estate objects).

Redirecting marketing effort allowed getting first significant results which were fixed through appropriately performed optimization.

Next step was to work on the site structure and interfaces: the way of presenting information was outdated (difficult navigation, impossibility to combine several buildings in one object).

Object list page

It was impossible to compare and sort out real estate objects by some characteristics except for prices in the old version:

The main element of the renewed website is the search system using the database of real estate objects. Searching is a tiring process, so we tried to facilitate this process for customers as much as possible, though preserving its compactness and flexibility.

Customers can see PIK Group realized projects and construction plans. Each object is accompanied by the information on events, mortgage opportunities, installments, construction stages, parking slots and so on. The outside infrastructure and transport roundabouts can be estimated in the map. In the long run it will be possible to show kindergartens, large shops, underground stations, and other infrastructure objects in the map. Everything mentioned above makes the process of the initial object esteem more convenient. The main advantage os such a realization is its visualization.

Most objects are equipped with web-cameras, so visitors can track all the construction stages even without leaving their houses.

After that A/B testing was carried out which resulted in altering some elements, namely, some of them were hidden under dropdown search box, others were put in the foreground:

An individual object page

The old design version presented all blocks of the same building separately, which strained the process of comparing and choosing: it was necessary to go back to the menu each time it was needed to look at flats in one block and start searching again:

The new page represents the detailed information: planning, booked flats and those available for buying, cost, mortgage interest rates in different banks, news about the object, construction history presented in photos and sales offices.

Main page

The main page was renovated alongside with renewing interfaces and design.

After changing the design critical and small errors were corrected: meta-tags generation algorithms for catalogue and other pages were developed and implemented, the site visitors’ behavior was analyzed, which helped take decisions on changing site details

In parallel, we deeply analyzed the topic demand to increase the attracted traffic and modernize the site for the factual demand. To do so, all the groups of needs were found and arranged by the number of their showing per month in the semantic demand for “new building flats”. Additionally, the non-covered or partially covered site traffic potential in the Moscow region was revealed, and the priority of working on them was proved by demand indicators.

The working process included the following stages:

1. Collection of search phrases from all sources accessible:
   
   • All queries and phrases from the current site structure.
   • PIK Group database of objects. The formulas to generate assortment matrices were made:
   By object names
   {Object name} buy
   {Object name}
   {Object name} a new building flat
   {Object name} a new building flat reviews
   {Object name} buy a flat
   Around MR cities
   buy a new building flat {city}
   buy a new building flat {city}
   buy a flat in a new building {city}
   selling new building flats {city}
   prices for new building flats {city}
   
   • Queries of site-competitors (uploading from specialized services, for example, SpyWords).
   • All queries accessible from site statistics counters.
   • MOAB database on key words regarding all kinds of phrases: “new building flats”.
   • WordStat Marker requests
   • Tips of search engines for marker requests.
   
2. Manual cleaning and clustering of the collected database containing search phrases. In the end, thanks to clustering there appear groups of requests which are actually needs.
3. Sorting out phrases by the total frequency descending.
4. Grouping and analyzing all needs, revealing actual frequency and global groups, detecting unused traffic potential for each search system per each need cluster.

Yandex

Google

The data obtained enabled us to receive and visually represent clusters with the unused traffic potential. By employing these general and more detailed data, we convinced customers to pay attention to concrete site pages: to give more extended information, develop separate pages (including the addition of new sections and subsections).

For instance, the section “Mortgage” required some serious development:

Also, we investigated the quality of working with applications in PIK Group and their competitors: a call back request was made using online-forms on sites:

We checked the time a customer waited to be answered, and the data obtained were used to make conclusions on the quality of conversion processing channels.

The time PIK Group consultants needed to answer an incoming call by the “hot line” telephone was also monitored:

In the course of our work on pik.ru site promotion, its visibility was in the range of 20-40%, which is a good level. But by the middle/end of 2014 its positions lowered, which was caused by Yandex algorithm change.

“Behavioral factors” became one of the most important ranking criteria, they can be formulated in the following way:

• How many actions does a customer perform on the site after the search?
• How often do customers find on the site what they are looking for (i.e. stop searching further)?

The leading portals contain dozens more real estate objects than those of developers, and they show prices. Therefore, they are far more useful for people than developers’ sites. Yandex can automatically analyze these things thanks to Yandex.Browser and Yandex.Metrics, so they can rank portals much better.

The main factors decreasing behavioral site criteria:

1. PIK Group doesn’t represent prices on concrete objects, whereas all developers considered to be their direct competitors post prices on their sites, for example, CM155, HBF-1, Morton, Etalon-Invest, LSR and Leader. But it was possible to get the concrete information about prices only by calling PIK Group sales office of.by directing all target queries only along one telephone channel, the customer reduced the conversion because of the following reasons:
   • The results of testing performed showed that feedback forms do not actually work: not all feedback queries were answered by call-center employees, and minimal response time through other ways was 1 hour, which was too long.
   • As for calling the “hot line” of the sales office the average waiting time before connecting an operator was about 2 minutes (including the answering machine duration), which was also quite long.
2. Imperfection (aging) of site interfaces. To improve behavioral characteristics of the site we offered prototypes of catalogue pages considering the major recommendation to raise the usability: present as much concrete information about flats as possible in a compact visual way.

Object list page

We provided the prototype with the mechanism allowing the traffic increase for low-frequent requests: tag- cloud at the bottom of the object list within which there should appear necessary catalogue categories:

1. X-room flats (“buy a three-room flat in Moscow”)
2. Flats at a definite underground station (“buy a flat at Perovo”)
3. Flats in a definite district (“buy a flat in Zamoskvorechie”)
4. Requests obtained at matching item 1 and items 2-4 (“Buy a two-room flat in Zamoskvorechie”)

On the one hand, it allows creating landing pages for low frequent requests, on the other, it won’t complicate the catalogue navigation.

An individual object page

The increase of information amount in pages did not presuppose any interface complication, vice versa, it was recommended to reject some filters (based on analyzing site visitors’ behavior) and intuitively unreadable icons in favor of text information and so on.