We receive dozens of requests here at www.opsway.com from entrepreneurs all over the world every month. All of them want to found a new ecommerce business or to improve an existing one. We have probably seen hundreds of launches, failures and successes by now. The reasons for success are different – a fluke or combination of the right place & time, an incredibly professional team or, say, a business that sells erotic goods. Reasons of the failure from our point of view in most cases are common – the irrationality of thinking.
Irrationality has a huge number of faces and behavioral patterns of course. A man running a small shop somewhere between Singapore and Australia thinks and acts as an owner of a multinational corporation. Launch should be ambitious and answer questions of multi-million dollar, and maybe even multi-billion dollar transactions. Someone who owns already successful and operating business wants to migrate to another platform, because s/he doesn’t like the current one – agree, it’s a very subjective metric for a decision making process. Developers often impose their decisions on business because they really want to play around with a cool new framework. Tens of thousands of years of developers are spent every month on creating solutions that do not meet market needs.
In this article we will try to describe cases in which we believe it makes sense to build an e-commerce project on open source platform. Moreover we will share with you the most common mistakes that we encounter daily on our market.
But to make the long story short, let’s deal with the SaaS question and answer it right away. We’re sure that many of you consider SaaS to be one of the most advanced and proper approaches to build an IT infrastructure for business. Shopify/BigCommerce + BrightPearl or Zoho and it shall be given to you. By the way, we also used to think so for quite a long time.
But let us be honest, the truth is that SaaS is just a way to do it quickly. It is not always about low price, and not always about what you really need. Yes, SaaS will cover 80% of small and medium e-commerce projects’ goals. But if you need some super custom logic or complex integrations, maybe even your own special warehouse processes to be built, once you need to work with huge amount of SKUs – it becomes difficult, and sometimes very expensive. Not to mention the problems you will face working in countries like India or China, most of the SaaSs are not tailored to their local legislation and you will have to modify everything greatly or use local solutions.
Therefore we made a conclusion for ourselves – SaaS is a great decision only on the MVP stage. If you move to the next stage, it is wise to take the proprietary or open source build and develop it. Yes, we know what you want to do right now – most likely you want to defend SaaS’s and tell us about your special relationships with them. You can become a key account for SaaS company and enjoy the custom features, premium support etc. But don’t you ever forget that your core system (and its functionality!) is not under your control, sometimes it can be very painful.
Open source platforms is a different story! They are free and on your servers! You can do whatever you want with them. A huge community will support you in all situations! There will always be a lot of people you can address your questions for free. Pfff, not in this world. The truth is that the decision to develop your IT infrastructure based on the open source platforms will cost you quite a lot in the future. It will be cheaper than buying licenses for enterprise platforms like Hybris, SalesForce Commerce, etc. but still. In any case, the cost of owning your e-commerce infrastructure and back-office is far from zero. Not speaking about its further development.
The decision to start development of e-commerce business based on open source platforms is primarily a financial decision. If you have outgrown the MVP stage, if you have a strategic business development plan for, say, 5 years, it makes sense to invest in your own infrastructure.
To begin with, let’s choose an e-commerce platform which will actually become the heart of your store. You can quite easily identify the market leaders if you have a look at the data from www.builtwith.com. There are a few open source platforms that can be quite an option for you: Woocommerce, Magento, OpenCart and Prestashop. In fact, you can use any of these solutions, they even say that WooCommerce works properly sometimes. At the same time you should think over and clearly declare your position on the market, you should also understand for yourself how much your business will grow in the future.
If you want to scale, if you plan & need to complicate the business processes and functionality of the store – it is reasonable to develop it on the basis of Magento 2. It most maturely addresses the issues of multisite & multi-currency, provides you with a huge number of marketing mechanisms and can work under high load with a help of existing on the market enhancements.
We must understand that Magento has two versions – Open Source (formerly Community) and Commerce (formerly Enterprise). Since our goal is to build an open source business – guess what version should we take? Right. Get ready to immediately add other solutions like Elastic Search to it, as it will be necessary for the store’s further development.
Once you have chosen Magento, welcome to the world of ready-made open source solutions – it’s time to hire developers and start building just what you need based on the given platform. The good news is that unlike custom solutions, in Magento projects it is enough to simply find or change developers/agency, thus your business is less dependent on the third-party companies.
This is when you will make your first record in the P&L document. The budget will of course depend on the country where you live. What you should expect is 4-5 months of work on average (team of 3 local developers or developers from more profitable regions such as Eastern Europe, India, Vietnam, etc.).
We do not know the amount of products you will have and where you will import them from. We also don’t know what amount of attributes there will be, your plan for building bundles and configurables, amount of languages you need, etc. But what we know for sure is that working with a catalog starting from say, 5-10K SKUs, will not be very easy if you don’t think about it in advance. Moreover you certainly will not manage it from the Magento admin panel.
Usually online stores use a terrible mix of Excel, external ERP data and some custom scripts. We suggest to do it in a more intelligent way using PIM (product information management system) solutions. Searching for the best solutions for our clients, we came across Akeneo, already mature project made by an excellent French team.
Another interesting PIM solution is PimCore, it can also be used as e-commerce engine if you wish. Unfortunately, we have not got a chance to try it out ourselves yet.
Integration of Magento and Akeneo, some other customizations will also increase your budget. But it will pay off in the future.
When business actively grows, at some point there are too many questions to be answered at a time. Warehouses, logistics, order management, accounting, purchase management, fulfillment management, CRM – who knows what will your business model needs to be implemented tomorrow?
Many e-commerce platforms declare the coverage of this functionality on one hand. On the other hand community with the paid extensions cheers up and it may seem to you that setting up all these processes inside an e-commerce engine is expedient. But this is a bad practice from our point of view. It leads to excessive customization of the administrative panel, which increases the likelihood of errors and affects the performance negatively. Moreover, if you decide to place your call center there too – it will definitely increase the overall load on your platform.
You should also never forget that at some point, for one reason or another you may want to change your e-commerce engine. If all of your operational processes are tied to it, it will be very, very painful.
It is wise to choose an ERP platform that you will make a core of your business. For most businesses back-office processes are fairly standard and thus there is no need to write them from scratch. We have seen and used a lot of different ERP systems, but in regard to open source ERPs – we did not find anything better than Odoo.
By the way there is an interesting project called Marello, which is built on the Oro platform, but so far we don’t consider it to be mature enough to be recommended to our clients. However we monitor the development and success of the Oro platform.
Odoo, which was previously called OpenERP, is a mature platform, it has many implementations in completely different business segments and can support you with qualitative development of all operational processes. The downside of the platform is the need to build a separate development team, because it is written in Python. However don’t you forget that Odoo will become a powerful platform for your business, not an obstacle on the way to your growth for many years ahead.
To sum up, we want to give you a piece of advice – do not make technical decisions on your own. Open source or not, when and at what stage it is worth implementing PIM, the need of a mobile version, which hosting to choose – all these questions require an assessment of people who work with them every day. It all depends greatly on the current technical infrastructure of your business, on your plans, budget and priorities. The rational choice can only be made by someone who has extensive knowledge and expertise, and intuition in your business domain proved by numerous cases.