biggest Russian transportation company
2019 revenue $0.7B
Business lines
Order form and tracking service for the
biggest Russian transportation company
Dellin.ru (Business Lines) is a Russian transportation company. According to the estimation of industry experts, in early 2019 a group of companies "Business Lines" occupied 20% of the market share in Russia that was about 0.9 B USD a year.
In 2012, I've created the main online order form, which was the first truly effective form on the market.
Moreover, I've built a package-tracking service, enclosing the web-communication cycle.
I am very proud that my solutions are still used and mostly unchanged.
Team
Artem Gendler (BA, Lead UX)
Tatyana Khryashcheva (UX)
Objective
Unite deliver order forms, calculation service and make it easy to use.
Time
3 months
Delivery is a complex process that requires a lot of information from customers.
The simplifying communication cycle is a key to B2C market.
Content
Context
My task was to analyze the current state of the order process if and ideate possible solution to reduce the amount and length of phone calls.
The existing amount of phone calls was too big and in case of scaling to B2C market this cause huge amount of expences.

Also, that was extremely important for acquiring new customers, especially from the B2C market because the average order from these customers doesn't have enough profit for a 20-minutes conversation.

On the other side, B2C customers don't have long-term contracts with us and it's very easy hor them to open another tab with other transportation services provider.
Discovery
At the beginning of 2013 company had a very blurry vision of online interaction with users. They understood that the current workflow isn't perfect but they didn't know how to deal with it.

My task was to analyze the current state of the order process. Understand where are friction points for users and find a way to solve these issues. At this point, I have a few conversations with:
  • Head of Internet projects
  • Head of Business analytics
  • Head of Call-centre
  • Head of Internal transportation service
  • Head of External transportation service
Research
I started with talking to people, after a couple of meetings with stakeholders, developers, and analytics I define the list of possible data sources for my research:
Web-site contact form
More than 15000 unclassified customer requests monthly.
Call center
More 500 call center operators worked those days.
Orders database
I was able to get the weekly extract of all orders, to analyze it.
Yandex.Metrika
Web analytics platform russian analog of Google.Analytics.
API
Quoting API and 6 delivery services API
The contact form on the web-site
At those time company had around 15 000 requests monthly, so research was statistically significant. This database contains the text of both request and response. After extracting all monthly requests I spend a couple of days reading them through and classifying them. Key findings were:
  • 40% of requests were related to parcel status
  • Users were unhappy with unpredictable price
  • Users don't understand verbiage on site
  • Users don't understand the difference between our transportations services
  • Users don't like that they need to enter the same information 3 times first to get a quota and 2 times to fill in the order form
Call center
I was able to talk to the call center lead and get statistics. Key findings were:
  • 80% of the overall call duration were related to the ordering process
  • Users don't understand verbiage on site
  • Users don't understand the difference between our transportations services and want the operator to explain it
  • Users don't understand how to fill in the order form and the operator provides the step-by-step guidance that takes about 20 minutes and is very expensive for the company
Orders database
Share of orders volume
Developers provide me one-week orders for me to analyze. Key findings were:
  • 40% of sendings contain one item and have dimensions less than 0,4m*0,4m*0,4m.
  • 60% of sending are delivered to our terminal as a destination point (users search for a terminal address to put it as a destination address
  • Users don't understand the difference between our transportations services and order Full-Truck Load far a small parcel
  • Users provide sizes like 22x10x8 meters but they mean centimeters
Yandex.Metrika (analog of Google.Analytics)
Collage usigng Yandex.Metrica form analytics data
Churn rate finding
Using an analytical system I found out that the in "Volume" field we have a churn rate of 19.62 %. Also at this service was there was interesting that if your cargo is oversized, you should provide information about that by yourself.

As an example I want to send iPhone 12 in package:

Step 1. Find and provide dimensions of item in meters from 89x165x28 mm you should convert to 0.089x0,165x0,028 m.
Step 2. Remember how to calculate the volume
Step 3. Multiply 0.089x0,165x0,028 = 0,00041118 m3 ...

Not so easy thing to do
Quoting API and 6 delivery services API
First, it was a surprise for me that typical package delivery has more than 50 parameters. Second, there were 6 different delivery services from letter delivery to interstate semi-trailer truck transportation. The third calculation service had very strict data limitations.

I started with analyzing requests from all to these services. You can see example of request below.
Example of ordering API request
I have analyzed 6 items like this to gain understanding.
requestType: cargo-single
delivery_type: 1
length: 0.3
width: 0.15
height: 0.05
sized_weight: 1
sized_volume: 0.01
max_length: 0.1
max_width: 0.1
max_height: 0.1
max_weight: 1.0
quantity: 1
total_weight: 1.00
total_volume: 0.01
...
There were 6 more requests like this. I analyze them in order to understand what's common between them.
Research key results
80 % of the overall call's duration were related to the ordering process
40% of the overall amount of calls were related to delivery status
40% of parcels amount contain 1 item
Version 1
It was a bit tricky but I found intersections between all order forms and calculations. Therefore, we get the ability to create a step-by-step order form with built-in calculations. So I create sketches and flow.

Bet before starting something big, leadership what us to prove them somehow that we understand what we are doing.

So as Version 1 I decided to slightly modify the existing form.

Below you can see the process of filling in proper data. One of the most significant changes was to make most of the fields disabled by default. This helps user to focus on one field.

Results were impressive, most of the users started to interact with fields continuously exactly as I planned and this reduced churn rate and improved conversion rate.
The team spends on development less than a week. But we have completed the following goals:
1. Focus users to fill out a form in the order we needed.
2. Help users with package volume (a lot of users had problems with this field)
3. Reduced churn rate and increased conversion rate
4. Reduced average time to complete from 01min 35 sec to 42 sec

At this point, it was confirmed that we can do what we planned and the impact of these actions will be significant. So we started the next step.
The team spends on development less than a week. But we new version shows significant improvements:
– Reduced average time to complete from 01:35 to 00:42
– Reduced churn rate - 15 %
Final solution
I created a solution that focuses on jobs that users want us to perform. In our case, it sounds like this: Deliver specific cargo from point A to point B.
The important thing is that only cargo parameters are fixed, A and B can be changed and sometimes users even don't know at this point exact addresses.

We simplify all choices and set as default most common options. So most of times the user doesn't have to make changes in transportation types. But if needed they are obvious:
1. Parcel (default) - most common choice
2. Full Truckload - this group contains 3 items:
2.1 Semi-trailer transportation
2.2 Van transportation
2.3 Hourly rent
3. Container

All other choices are simple for users and easy to fill in, as an example parcel dimensions.
Next important thing while filling in, users gets a real-time quotation. This affects conversion and sales of additional options. We calculate each and every option so users will know how it affects the price.
Design
All goes well on previous steps even coridor testing. So at this point I get additional help from Tatiana to create all final design. Those days we used Photoshop for that, but I've recreated it in Figma.
Results
+ 15 % conversion
Simple and strict workflow delivers users efficiency they didn't have any time before on the transportation market.
- 20% support requsets
All limitations were clearly explained for users so they didn't became frustrated with this behaviour.
Check Online Order
in Russian language
Made on
Tilda