LUN.misto — urban space research and quality of life assessment

Product
LUN.misto
Category
News, Statistic
Role
Product
Year
2021
The project is very versatile, combining several types of services, from an air pollution map to articles about home improvement/research and real estate prices.
Problem
Point solutions to offer improvements to the product site with minimal programmer effort.
Solution
Look for problems in the analytics, make hypotheses, and offer simple and quick options to implement them.
The process
Writing questions and briefing the client
Writing questions and briefing the client
«Dives in and finds out everything in 1 day and subscribes to Telegram. Sees and immediately writes appeals to 1551, asking the authorities — why bad with landscaping, submits a project for the city budget, such urban-activitis, which uses the data to change things and the link throws in the personal Klitschko»
The perfect flow
Important pages
According to the client (what they aim to develop, what they see as scaling), according to the interest of the audience (analytics, social networks).
Got access to Google Analytics of the project and started working on hypotheses, confirming or refuting them.
Map
Prices
Pulse
Research
Hypotheses
1/6. Low page load speeds
More than half of all users use Chrome
Average page load speed for al browsers
7.4 sec
Average page load speed on Chrome
14.3 sec
Desktop users have problems with loading speed. Mobile users have problems with the home page.
Check out the PageSpeed Insights website for speed estimates:
Home
Map
Solution:
perform photo compression;
use JPEG 2000 and JPEG XR (they compress more efficiently than PNG and JPEG).
How to check? We'll keep an eye on:
average download speed;
bouncesrate on «Home» and «Map» pages.
2/6. Ugly Duckling?
86.8%
of users are leaving the «Pulse rinku novobudov» page
79.5%
bounce rate of Pulse pages
23.4%
of users scroll 10% of pages
Going back to the home page users look at stats in other cities
Solution:
changing the city on the Heartbeat page;
give an opportunity to go to a similar section with «Monthly and quarterly statistics» statistics;
increase the value of the data, for example by comparing with another period.
How to check? Let's compare the data for the old and new period:
How to check? Let's compare the data for the old and new period:
bounce rate;
scrolling statistics on pages with Price Pulse.
3/6. Where are the articles?
Our user has to suffer to read the articles.
22.5%
of new users
22%
of all don't scroll any further than the new building statistics
Percentages of clicks on the most popular articles
Solution 1/2:
a lot of fields, texts and illustrations;
How to check?
let's see the clicks on the menu;
let's check which naming is better with A/B tests.
Solution 2/2:
add a link to the articles on «Map» as the most visited page.
How to check?
let's look at the number of clicks on articles from Maps:
let's track the depth of scrolling articles, views, and number of phers.
And it works best:
advertising in google search (the bounce from google is the least)
referral links from specialized sites about life in the city, such as https://bzh.life.
When there is interactivity and a lot of infographics in the articles, they spend more time and come back
4/6. What about Exit Rate?
Infographics without links at the end.
75%
Exits
25%
Next Page
Article with links at the end.
66%
Exits
34%
Next Page
83%
of users don't get to the recommendation block
Solution:
the "recommendations" block;
links in articles.
How to check?
see if pageviews increase;
let's see if the percentage of Exits has decreased.
5/6. Subscribe to Telegram
11%-14%
of users click on «Subscribe to Telegram»
More explicit CTAs may be the reason why more people go to TG from the Home or the page «results2020»
12%-18%
of users reach where the «Subscribe to Telegram» button is
54%
that click — come from Google
27%
that click — come directly
2%
that click — come from Facebook
Solution:
place a button with a link to Telegram at the beginning of the article
place a Facebook button
How it was
How it became
How to check?
let's check the number of clicks (in %) on the «Subscribe to Telegram» button and compare it to previous period.
6/6. Hard to find and easy to lose price statistics
There are no direct links from the Map to the Price Statistics. Where do they come from?
This way goes 0.25% of users
Of these, only 22.3% are new
Does LUN.misto create demand for quality housing among new users?
Few clicks from the main page to the Price Statistics, which contains a link to this article 2.82% of users go this way
What if we put an emphasis on price statistics on the Main Page and then see if it affects the demand for housing?
Solution:
put a link to the Price Statistics on the map;
emphasize the block with prices, raise it higher than the Market Pulse.
How to check? Check the Price Statistics Conversion:
from map of new users;
from the home page of all users.
Credits
Iryna Stryhunova
Antonina Cherenkova
Maksym Chervynskyi
Julie A
Anastasia Tychynska
Victoria Gumeniuk