Creating a cutting-edge ERP/CRM platform: How car service stations conquered the sky
A Canadian body shop enterprise owner and our client in combination (let’s name her S.) provides auto repair software. Being on the market for over 30 years she could no longer stand the current level of the body shop system and came to Belitsoft with an idea.
It was an intention to create a software that surpasses any other system ever existed anywhere in Canada. And just because car service entrepreneurs used something similar to fossil Excel, it seemed simple. But not so fast.
The problem we were targeting to was the lack of flexibility. The brand new product was expected to be simple, versatile and yet affordable.
The main requirement was to develop a complex software that is capable of performing as MSO (multi shop operator). But still combined with a few more details.
This miraculous body shop software was supposed to have… well, has now:
- Functions of ERP systems, and a CRM;
- SaaS (Software as a Service) model;
- It integrated all the systems and software habitual for Canada or required for body shops to operate effectively, respond quickly and be permanently available for the customer’s side;
- It is multi featured and affordable+scalable both for large body shops with branches scattered all over Canada and small local ones.
How have we made this? Let’s see.
From 2k15 up till now. The team: does the size matter?
The story begins back in 2015 and still goes on. Many brave and faithful warriors were indulgently cut off without the slightest loyalty of the queen... But who loves to start the movie from the middle?
There were three of them: a Chinese duo and a Canadian lady. However, soon the plate was flooded with wrong ambitions and she remained alone. Beautiful and sharp as a sword, S. with her followers took the reins of the project.
Since the August 2015 till January of 2016, the development had been done by the one and only Andrew Primus. But for how long could he hold the project? Nobody knows. However, before it all gone out of control it was agreed to find more people to translate the queen’s vision into reality. Since then the team has expanded rapidly.
In December Alexander Secundus and Eugene Tertius were interviewed and successfully hired. From that day on, the future of the project was on the shoulders of Atlas split into 3 human beings.
The deadline was set: 8 months for viable results. At this point, the count of three still wasn’t enough. Summer welcomed another two developers which shared the fate of the team until the harvest season swept away.
2k16. “If the landslide isn't stopped, it will bury real jazz.”
In October the project, at last, snagged a tester. The team greets Michael Viscum.
Closer to Winter the clocks went backwards. From November to December we’ve lost two of the team. In August of 2016, Alexander Secundus, one of the mighty old-timers, laid down his weapons leaving the total number of three troops.
The queen had a counselor, Peabody, who was a supportive link between the customers, the staff, and the queen. He worked nights for over a half a year and suddenly decided to try himself at something new.
Peabody asked developers for tasks but was rejected. So he made his own. The queen had some expectations of all the action, however… Royal hopes soon have shrunk. The development straw didn’t fit the counselor’s mouth and just sucked him up (BelItsoft guys had to clean up all the mess left afterward).
2k17. The year with a smell of curry
September of 2016 brought life to the fading team. The need for a mobile app led us to the forest nation where we pulled out two brilliant mobile developers: Nikita Pix and Eugene Vir. These elves made the excellent job and disappeared before the sunlight touched Himalayan peaks.
The created mobile application is totally usable all across Canada, provides a full range of services from BodyshopConnect, takes photos and sends the info directly to the users’ accounts.
October 2017. Her highness S. asked for a tech support in the modern lands of White Ruthenia, Belarus. In January two Snow Maidens, the representatives of English and French sides appeared. However, their presence fades away in Spring.
The feeling of “mistesting” was hanging around. So, in April of 2018 the team grew by another tester Sergei. He was joined by two more support ladies and a missing developer.
Cheap & numerous vs European quality. At some moment S. realized that she needs a cover-up. Cheap and fast but with the same benefits as us. Guys from Pune agreed to help and attend the project development. We embraced our differences and were ready for the challenge. Well, it was the longest three weeks in history…
First of all, those guys worked in shifts. Eight hours of the day a person worked on the task for S. Not a minute longer - there is another shift waiting to use the computer. 24/7 development - thumbs up! Exciting, huh?
Secondly, the results were riddled with mistakes. Our guys had to postpone their own tasks to fix the newcomers’. It naturally increased the total development time. S. once asked why Andrew started doing less. All we had to do is to show how much he reviewed and fixed.
In two days, when the results finally reached our tester, Michael, the volcano erupted. He literally overdosed, so in 7 minutes all the tasks done by our “helpers” were returned for fixing. Again.
It was a special form of delight to behold this circus. However, the queen didn’t share our enthusiasm of seeing how far the degree of idiocy may go. She picked up the jaw and kicked those dunces off the project.
The ambush in the shelter.
The middleman problem. The conversation with S., the queen, hadn’t been held directly. As was told before, she had a counselor who did his best to interpret the royal preferences. However, the measurements of the “best” start fading once you realize how poor the carrier is. The original plan wasn’t thought out: human factor crept in and the dearest counselor disappeared in July of 2016.
At the time of Peabody’s exile, Michael becomes fluent in royal language so he talks the talk when the queen needs the info, and the team is seeking for a feedback. Everyone is satisfied.
The internal interaction. Pinging IT guys is a “great idea” that visits many customers: “How long? Oh, you can do it faster!”. In such cases, the team can wish you to “Do it yourself”, sprout some wings and flutter away.
We overcame this showing the reason for the time frames. The queen once visited prospective clientele and decided to show them the freshest we had. She wanted to do it on Tuesday. We were scheduled to release on Thursday.
Developers warned her that the code is raw and nothing is going to work. However, as it always happens to royal family, warnings have all gone unheard. As the result, the demonstration flopped and she had to apologize to both clients and the team. The lesson is this: heed the conclusions.
Your budget and project scales. Carefully planned budget is important. But it should be flexible. IT project is a mix of the original request, customer’s vision, developers’ capabilities, and ongoing preferences and adjustments. Easy to see that the last aspect is spreading far beyond the original plan.
Some time afterwards your project may expand. The growth of ambitions should be backed by the financial support. Otherwise, the progress is stuck.
You’re lucky if you had a chance to grab people who are ready to learn the extra skills. For this project, our team became a bunch of versatile pros that together pull the whole project up.
Project birth. The notes of the midwife.
“That was a tough 9-months period with over two years of adjustments and testing…”
This system itself is a multi-layered mountain. On top we have label functions (franchise); MSO on the slopes, and single body shops in the bottom.
The question was where to start this mountain.
If you want something to be done right, do it yourself.
The ways are: from the top to the bottom or vice versa. The difference is that if starting from labels, it’d be easier to think through details and MSO finer points, and the bottom will be done “in seconds”. But, we should have been ready for years of constant development before the final release.
“That was a tough 9-months period with over two years of adjustments and testing…”
On the other hand, “the bottom up” approach allows selling the results in months. Getting feedback and money in several months looks much prettier. The pitfall is that the troubles with MSO and other more complicated elements from the top appear as soon as you complete the bottom. And because the peak didn’t exist at the time, you haven’t even thought about it.
So, what was chosen at the beginning:
When it become clear that several years with nothing in hands don’t satisfy the royal heads, the development gone from the bottom up. After it was done and the team was ready to conquer the middle slopes, they faced challenges.
It was the time when the queen shared the throne. The counselor made his best trying to interpret everything S. wanted. But humans are humans with their own way of thought. So, Peabody rephrased the royal vision to the ground leaving only a few blue-green brushstrokes of the original “Ninth wave”.
The deadline was clear - 8 months. We managed to create everything the counselor described. And the moment our queen looked at the results, she realized how she screwed up.
“Peabody’s creation” was plain, poor and didn’t work at all. Chinese whispers did the job: everything went head over heels. And that’s where the “sweat & beauty” phase of birth started.
Seeing how 8 months were blown in the Black Hole of Nothing, the queen was blood and thunder. She took the control of the situation and changed the approach. So, we had an extra time to fix everything according to the direct conversations with S.
Because we wanted to have something to offer promptly, the new approach worked. It took us a few months to release a demo version and give it to people for a try. The team was blessed with the constant feedback and suggestions from the first users.
As Michael said, “The feedback you get from the real workers [end users] is more valuable comparing to the testers and developers’ who don’t fully understand the scope of application.”
In the end, alongside the development we had a flow of requirements from the customers and additions from S.
The underbelly of the beast
First, ongoing requirements and adjustments delayed the finish line taking more time to develop.
Next. With the “bottom-to-the-top” approach we faced the problem: the relations between the inner elements of the system became too close that led to the co-dependency. Any adjustments in one place would impact the other links of the chain.
This blunder is still on the menu, though not as pronounced as it used to be in the beginning.
“I need it yesterday”. Good way of motivation that could cost you quality. Software development, especially of complex and extensive platforms with ongoing support, requires time and good specialists on top of your deep understanding of what should be done.
Yes, you have to learn how to tell what you want. But the pitfall is deeper. After you find a good team and gave them a proper guide into the future product, the next step will be to hurry. You want a reward.
You’ve made 10 steps towards the goal, so why doesn’t the goal make its own 10, right?
In a rush, you are not able to explain your vision properly and listen to the response from the team. Skills you pay for matter, save a great deal of time and give you peace of mind. But remember the importance of communication with the ones who translate your ideas into reality.
S. trusted the communication to Peabody and it cost her 8 months and additional time for us to fix everything.
Multi Shop Operator: another stairway to Heaven and a highway to Hell.
MSO is a piece of a management system that is used to subordinate several body shops under the one corporate guidance. In short, it allows a fast spreading of changes in the whole network of body shops simultaneously, tracking stats, quality and income etc.
It covers body shop networks of unlimited sizes, synchronize their management routine but leaves a freedom to perform individually.
The trick is that Multi Shop Operator is a unique piece of software. So, the team had to develop it from scratch basing on the queen’s vision alone (that was ludicrously misreported at the beginning). And because the start was from the bottom up many aspects weren’t thought out properly.
However, based on the feedback and development process issues, we solved the quiz in a few months so now it’s widely used.
Peabody with the chinese whispers are gone. Since S. took the reins the real work started.
Structure. Every week our team receives emails with an extensive tasks list, detailed instructions given individually to each developer, and the additional information about new adjustments.
Regularity. During our 1-week sprints, we have 3 precious overlapping working hours in touch with S. At these times our boys left office at 8 PM while S. and her team came to see what’s going on at 9 AM in Canada. Well, 180 minutes never were as fruitful as those.
Testing. S. surprises the team with new ideas like a wind down in the subterranean caves of Afghanistan. So, in search of stability (better life), our team used manual and automated testing. Well, that feels way better: now we prevent even the slightest change from turning into a weak spot. Moreover, the testing allows S. to track the news about her creation.
The Team. After everything was settled down, the team started to dwindle. And, to calmly meet the budget, Michael and Andrew were asked to fill the gaps. Michael, QA tester, became a tech support lead and project manager. Andrew, like Neo from “Matrix”, catches up and implements tasks with the speed of a bullet, plus helps novices to get in this mix.
Now (with all respect) they may be called Genetically Modified Swiss Army Knives, just because of the resulting versatility.
2k18. Ready for the growth?
Time passed and the software value moves on upwards:
- 3rd party apps integration. The system turned into a huge blanket that covers multiple systems and gathers modern software Canada is accustomed to.
- A support team from belarusian side provides services in both French and English.
- Current integration supports around 12 different systems. Including Cloud accounts, Google etc., so that the info exchange goes seamlessly.
- The repair process is fully automated and does not require a direct presence.
- In the mobile version, you choose the body shop, the type of service, take photos of the damage and upload them to your own account.
- The system is almost everywhere, gradually conquering the hearts of Canadian body shops. At the moment it handles over 200 body shops, including more than 40 from the CSN network.