Case Study: Successes and Slip-ups of Agri Trading Platform Development0
- 6 June 23:46
Every agri tycoon wants an agri platform to easily access the data, raise capital, or even sell the platform. Almost everyone tried, but few succeeded. Many of them had a negative experience in the past, others witnessed failed attempts of the other companies or found the competition too keen.
Despite all the reasons why businesses avoid launching an agri trading platform, we’ve decided to enter the market with a brand new solution though.
What Agri Platforms Are Used For?
The goal of an agri trading platform is to optimize the business and routine processes through digital technologies. And there are many reasons to launch one based on the benefits it can bring.
- Farmers can get the pricing details first-hand as soon as the offer is created and understand the market price just from looking at the trading tables.
- Farmers see what products are in high demand on the market hence more profitable.
- Understanding what price offer is optimal, farmers can promptly sell their products.
- Identifying the best price traders can negotiate better deals (e.g. farmers can close the deal by supplying the same dry freighter with their production and send it to export right away).
- The traders can get the advantage to see product offers in advance before the rest of the platform users can do so (although NMGlobal doesn’t have this feature, it still can be implemented to any platform).
- The platform may be an extra source of income if offered for use to other agribusiness owners and traders.
The Motivation Behind Wanting to Launch a Platform
There is a number of reasons why the entities involved in the agricultural sphere would want to launch a trading platform. However, we would highlight the key factors:
- Ambition. Growers and manufacturers work hard and love the process of making something out of nothing (million-dollar company from small business, hectares of fields with wheat or other grains from a small seed, etc.). Driving the complex mechanisms in order to get the needed product, they want others to associate it with their names.
- Next-gen innovations. In most cases, farming is a family business where a rising generation has been involved as young as 8-10 years old. Working from childhood and watching the whole process they grow up and want to automate the business. They evolve innovations to optimize expenses and increase the revenue (e.g. transferring paper documentation to online).
- Trading optimization. Agribusiness owners desire to optimize the process of trading and negotiating deals. They need a resource for that displaying the average price for a specific product at a particular moment or an estimate for the future week. Plus, a trading platform can simplify negotiations, so more deals can be sealed.
Why So Many Fail?
Launching a platform people working on it, in many cases, fail after release or get stuck in early stages without even presenting the product on the market. So, why are there so many fails?
- Lack of understanding of the agricultural business by those in charge of the platform development (often, agencies or external consultants). The agricultural sphere involves many different entities including farmers, traders, logistics, brokerage, cargo handling, and others. Which you would only know about if you studied the industry (e.g., by interviewing the people involved, analyzing their routine). Every single entity interacts with the rest under certain rules. And if you have no idea how this communication works, you can hardly offer benefits to the entities.
- Offering useless tools. Without a clear understanding of how the entities operate (their internal environments), not analyzing and getting into the details of their work routine, the developers make up false processes, try to copy the other platforms applying minor changes, hence create useless, inconvenient tools that none of them needed in the first place.
- Forcing users to do extra work versus adapting to their existing processes. When the platform requires extra efforts, retraining for being ready for a new work scheme, unclear values, it is rejected. For instance, the traders already have trading partners, obviously, they will not understand how they can benefit from using a brand new trading platform.
- Choosing the wrong investors. Opting for investors who don’t have a clear understanding of agribusiness and how it works, who don’t aim to invest into a product that would work (not giving the corresponding task to investigate the product) usually results in failure.
- Focusing on a single entity. Some platforms enter the market offering narrow orientation on a single entity which makes them inconvenient to use for the rest of agribusiness participants. Lacking proper communication with the other parties that should be involved in the business, those using platform will also leave.
- Failing to find a good approach to SEO. Agri trading platform is a complex digital product including tens of thousands of pages (with product categories multiplied by all possible locations). Such a platform requires an automatic system that would: 1) monitor page positioning in Google; 2) advise on actions to get the pages to the top; 3) decide on what pages to promote first according to business needs.
- Lack of offline promo activities. Farmers don’t usually surf the Internet to research websites, platforms, and other useful sources. To make them use the platform one has to combine digital space and offline and hire a salesperson who would visit farmers and traders separately with a product presentation.
In short, managing the development of the agri trading platform is a comprehensive process that needs experts with both a systematic and creative approach. A creative person will find out how to make the platform stand out in a crowd while systemic one will build a financial model writing a sales script for every single entity. You will also need a person who will manage the sales and a team lead to manage the developers. Read further to find out about the specialists you will need for your project team.
Working on the NMGlobal Project
We met our client at a business conference. The next step was to find professionals who work with this kind of digital projects. And once we did, we carried on negotiations within a month and ended up signing a contract. Ahead of us, there were several stages including research, technical task, design, development, and actual launch.
So, let’s start with the research stage and then move further on the rest.
01 Research Stage
The research stage is critical when it comes to creating a product that would satisfy users’ pain and gain since it allowed us to:
- Study and analyze existing platforms — their difference and value proposition, the way they promote, what they are used for, etc.
- Study how the agribusiness works — to understand the rules and processes of this industry, communication between the entities.
- Define our unique value proposition that would distinguish our platform from competitors.
To get the data necessary for that and for creating a prototype, we did the following.
Conducted deep interviews
We conducted a series of deep interviews with the people involved in the agricultural business, including more than 30 interviews with different farmers. Among other valuable intel, this gave us an understanding of how they work, start their day, what they do all day long, what they search for on the Internet, how they use platforms, etc. We put ourselves in their shoes, so to speak, to understand their routine and bring out a helpful product that would optimize this routine.
Processed the interviews
We noted down all the key information after each interview which helped us create user descriptions — sort of a portrait of their daily schedule that shows how they could benefit from a platform. Next we (the team consisting of a product owner, UX designer, and system architect):
- Compared the user descriptions and isolated data into three categories: systemic, non-systemic, and traditions. In other words, what is consistent in their routines and what doesn’t influence the end result;
- Identified where each entity experienced losses to better understand and recognize users’ problems;
- Reviewed the interviews and determined key information. We identified the entities involved, their ultimate goals, risks, and processes, how to call the entities; what filters to apply to capture everything related to production (e.g. humidity, tonnage, grade, level, etc.)
- Created a prototype in Photoshop.
What we discovered during the interviews
The interviews gave us some valuable insights into the agribusiness and how a trading platform can automate its processes, saving time and allowing to trade multiple products simultaneously.
- Just how complex the trading process is.
Any business task is. But this was about learning what actions go into growing and selling products. Here are two examples and how a platform can help:
- When farmers sell products, they offer a price and get multiple calls from traders offering their own price (it could be from 10-20 calls). Then, the farmers give traders a call back with counter offers conducting a kind of auction. The platform automates the process allowing to do all this online without the calls and even trading multiple items at the same time.
- Traders have to visit at least 5-6 sources to see logistics cost, the cost of a grain elevator, and what price the other traders offer for certain produce. All these processes are transferred onto our platform saving time to all entities.
- Big farmers and traders lack the tools to effectively track employees’ productivity. In most cases, they use IP telephony and screenshots which aren’t that effective for monitoring employees productivity. The platform offers multi-user control allowing to keep track of KPI (e.g. how many calls and bargains an employee has made).
- Farmers need digital products that are easily accessible from a smartphone. Since farmers deal with many issues on a daily basis, they stick to smartphones and prefer simple technologies. Our platform is mobile-friendly and adapts to smartphone screen with ease.
02 Technical Task
The technical task doesn’t determine the way the final product looks like. It is a skeleton and guiding line of the project that shows us what needs to be done.
So, let us showcase the features we added when working on the tech task.
- Calculator for farmers. Allows calculating the cost for storing, grain cleaning, elevator services; considering physical indicators that influence the quality of produce, like weeds and humidity).
Inputting features of products to calculate the cost of an elevator.
- A map with nearby elevators. Allows computing the distance from the farm to the nearest elevator as well as the cost of transportation, storage, and drying).
Listings view with a map of elevators and route from your location to the nearest one.
- Table view for traders and Telegram integration (in the near future). Traders use Excel to change the price every day, make screenshots of tables and share them on a Telegram channel. Our platform has the functionality to automatically post to Telegram or Viber once a trader enters or changes the price on the platform. This is going to be a great indirect ad for other traders in a Telegram channel.
- Mobile version. Any party of agribusiness has a mobile device at hand. Thus, we adapted the platform to mobile devices hence making it more user-friendly.
Mobile view of the platform.
A senior designer was present in all the meetings, interviews, and discussions. He also needed to understand the entities to aid in prototyping, so he listened carefully, asked probing questions, and made notes. Together with the designer’s professional competence and experience in creating a convenient interface helped reduce the number of edits and resulted in higher quality.
Creating a listing that includes the type of produce, culture, characteristics, weight, and price.
We also implemented solutions to simplify navigation for NMGlobal platform users. Now it has:
- Integrated table view. This feature will let farmers and traders view, add, change price, etc., like in the Excel tables they’re used to but with less hassle.
- Integration with Telegram. What’s manual in the existing process of the traders’ daily schedule becomes automatic.
Technologies we used for building the platform
Since we already had experience developing complex digital products, we decided to use the same advanced technologies. We went for the high-performance React for the front-end. It easily handles large volumes of data, traffic, and complex operations. React guarantees continuous support, better SEO, fast loading, and responsive UI.
As for the backend, we preferred PHP to Python. While comparing these two technologies, PHP is more cost-effective than Python in terms of changes, bug fixes, improvements, and developer support (hiring Python experts for every update can get expensive).
When it comes to selecting the appropriate technology for a project, we don’t recommend chasing trendy tech. There’s no guarantee our clients will find developers to maintain the platform once all work from our end is completed. See what technology is more beneficial from the financial standpoint yet satisfies all your goals — more affordable one doesn’t mean cheap.
Who Worked on the Project
When considering the team needed to build such a platform, we can highlight the following participants:
This position requires critical thinking and creative idea generation. The product owner should be in design like fish to water, have experience launching products (hundreds or thousands of products, from small to big), understand product cycles, and be able to criticize the team. The product owner can be in-house which will save investors some money, help the product owner to feel connected with the project (evolves together with the project, takes pride in the work), keep the data about the project and pass it on to the new employees.
This is the person responsible for the visual representation of the project resulting in the future interactions between the user and the platform. They make sure that the platform design and interface are convenient for use. We had our designer present on all meetings and interviews to save time when prototyping. It’s not recommended to change the UX designer in the middle of a project, otherwise, you will waste time introducing the processes to a new specialist.
A professional who sees the whole architecture right from the inside of the project. And makes sure there are no issues with the logic of the software or other inconsistencies.
Project Manager (PM)
The PM understands both the design and programming, manages the tasks schedule and delegates the tasks to the right teams.
They coordinate the work within the respective teams (front-end, back-end, etc.) and see to it that tasks are completed and deadlines are observed.
And, of course, you need people who will write the code, the functionality of the platform and its interface, fix bugs and deploy the software.
How We Selected the Team for the Project
Getting specialists from an agency gives you some major advantages right from the start. At the very least, you have a wider choice of professionals when recruiting for our project. You also hire people who fit the criteria:
- Competence (what projects, functions, features they worked on; how easily they pick on tasks you discuss, show a creative approach to solving them).
- Involvement in the project (whether they are proactive and ready to offer new ideas and ways to resolve the issues; whether they can think outside of the box and offer you fresh solutions to problems you present them with).
05 The Launch
We launched a unique platform that doesn’t have alternatives yet. There were some major factors that helped us in the process.
Factors That Led to Success
- Created a platform (not a simple website) that brings solutions for different entities and solves their business problems.
- Gathered a great team of high-skilled professionals working on the project (UX designer, system architect, managers who worked on researching the target audience and conducted deep interviews).
- Built the logic of the platform on the knowledge and information received from the interviews.
- Kept on optimizing the already existing processes instead of creating new ones (we didn’t try to reinvent the wheel hence saved time, money, and effort).
- Kept on A/B testing throughout the project starting from the platform prototypes to its design (to see how entities react to each element of the platform).
Challenges We Faced
Without a doubt, everyone experiences challenge while working on different kinds of projects. Agribusiness is a complex, sometimes even tricky, area and demands to consider many aspects. Thus, we also caught some bugs during and after production:
- The produce requires categories and subcategories. Let’s take honey for instance. It has a number of different cultures which should be mentioned in listings. However, it slipped our attention when we started the development.
Selecting culture when creating a listing.
- When a user receives a new message it should go to the top of messages sidebar. Another slip-up, one you don’t see until the system is ready and you can test it.
The messaging system.
- Whether a user receives a request for collaboration, the next specific action is missing here. It’s a bit confusing where should clicking on the request lead the user, to chat or to the other user’s profile.
Notifications about collaboration requests.
We came to the conclusion that a bigger team make fewer mistakes during the development process. Two heads are better than one when you need attention to details.
Additionally, you can locate shortcomings only after a user actually tries a feature and can give you real feedback. In our practice, when farmers tried the calculator, they asked to add the optional field for the cost of storing per ton of produce.
How to Promote a Platform Before the Launch?
We were interested in making the platform known to as many people as possible and acted correspondingly. Here are a few recommendations on how to spread the word about the project prior to its launch:
- Prepare promo-clips while the project is still at the design and prototyping stage.
- Monitor conferences, expos, and events where you can promote your product for a year in advance.
- Create a social media channel while still at the design stage to communicate with potential clients. As a result, you’ll get a more loyal audience and a ready platform for advertising.
- Advertise channels by means of YouTube, AdWords, and SEO (in our case: over 100k followers on YouTube in a year; indexation by Google in 4 months; PPC advertising expected in May).
Advice We Could’ve Used Before Working on the Platform
- Delegate writing of the technical task to the back-end team lead. They see the logic of the end product and interactions between its elements and can prevent mistakes throughout the project.
- Hire an independent QA. Your QAs are too involved in the project, they may not see certain bugs. Meanwhile, a third-party opinion can highlight them.
- Hire top-level PMs. Training novice project managers (seduced by their low prices and high motivation) may cost you extra money and time on fixing bugs in the result. Place the bet on top-level professionals if you want to scale dynamically.
We developed a platform for the European market as well as developing markets with minimal competition. The decision was to first enter the Ukrainian market since there is low competition among platforms, while the competition between agri companies is unfair (a peculiarity of the Ukrainian market). So we started here, next will be Poland and Scandinavian countries.
When developing a platform that would cover a wide target audience as possible, it is first necessary to understand the main goal. Our objectives for the future is to:
- Become #1 in the segment in the local markets that we target.
- Enter the international market (Poland and Scandinavian countries).
- Become an exchange, so to speak, and guarantee compensation in case of a failed deal.
- Make a fortune.
Additionally, for anyone planning to create a similar platform we’d recommend:
- Using Scrum.
- Choosing a team management tool (e.g. Asana).
- A/B testing everything (prototypes, features, entities).
- Using only actual data, not opinions.
Previously, we wrote that “If consumers do not see anything unique in your brand, you have consumer goods” span>
Did you like this article?
It seems like there should be enough information on the...96
Being an owner of a small local beauty salon is...0
Building a website is half the work. Keeping it afloat...0