A not-so-recent question on Quora asked whether custom (=built from scratch) solutions for CRM are better than the standard products. The question kind of tried to answer itself by pointing to a comparison of custom vs. “non-custom” system. In summary, the analysis compares what it means to build a custom system compared against salesforce.com professional edition. I am sure the developers/promoters had all the right things in perspective within their capacity. But, in this age of one-click installs do we need custom CRM? You must be joking.
The question is old, is self-promotion and had answered itself, but for nothing but curiosity I wrote:
I know that the answer is within the question. But custom CRM solutions are so passe today. Except for shops that do not have any solution that remotely addresses their business needs, custom solutions only add to the implementation complexity (many times business does not really know what they want), and maintenance complexity (add tonnes of functionality that exponentially increase cost).
Yes, micro-niche sites do not have packaged CRM solutions. Even for the so called ‘standard’ industries, any solution available out of the box is not a straight fit. For e.g. Zoho CRM provides all entities needed by a typical sales org. But each sales org is unique, and that induces the need of customizing the software. Does it mean that I stay away from the product market altogether? Does it mean that I will trust a development team to build ‘from scratch’ solution based on PHP, .NET? Seems quite a stretch for this day and age. Packaged software for a typical CRM buyer provides a better baseline as compared to custom one, and that makes a lot of difference in driving the success of the end product.
I expected this to die down as most of the questions on Quora do, but was surprised to get a response:
I would actually disagree because a development team with the experience can correctly implement a solution by asking the right questions. Drupal is a common platform and many businesses are using it today because it allows for so much customization. Sometimes it is just knowing how to correctly implement it. Businesses often do not know what they want, which is true, but if the right team is on the project, they can get exactly what they need.
Again, let me reiterate. I am a developer, and do understand what a good development team can do. So I do agree with the poster that in the right hands the custom solution can become a God-send.
However, businesses are moving away from custom solutions because good solutions with a palatable risk are hard to find. Unless you are doing something very specific that is not covered remotely by the existing CRM systems, I wonder why would anyone want to go ahead with this. So, went my response..
Thanks for your comment.
1. I often end up comparing instances like this to ‘business core’ solutions like ERP. Very few people think about implementing Supply Chain, or Financial Systems from scratch. The complexity that people encounter by thinking about business problems today, and the way it can change tomorrow, can be dizzying for a medium to large scale business. If I apply that argument in CRM, it will just be a matter of which option makes more things easier to implement, easier to maintain, and guarantees the scale (in terms of # people, future business expansion, optimizing processes and the like).
2. CRM processes within an industry stay more or less the same at a high level, while getting into nuances and complexities for micro-verticals and non-standard industries. So then, copying a standard process that has enabled industries of the same size as me, or larger, will make things easier as compared to any re-invention. Out-of-box software is a representation of those processes.
3. While skilled set of developers can easily mimic this and go about copying frameworks across their implementations, they can never reach the same scale as software makers. If they do get the formula right, they will be looking at putting it to use as a product rather than starting from scratch.
4. When there is a platform (e.g. Zoho) that I can get started today and get my processes in order, there is little incentive to go the custom way. The incentive in reduced license fees have often been seen gobbled up in higher maintenance costs (consider enhancements + fixes)
5. At the end of the day, there are custom implementations which have migrated to standard CRM software because of on-going issues. There are other custom implementations that are running for more than a decade quite optimally. Both work, but clearly custom implementations introduce a lot of risk for the business in getting the things right and keeping them right. Skilled developers help (a lot) and can add tremendous value, but that will clearly not be the chosen option when any of the existing platforms are considered “good enough”.
This topic has been close to my heart for various reasons. In my experience I have started out as a strong believer in custom systems that are made-to-order for the client. But gradually I began to see no points there for business to follow that model. As mentioned earlier, humans are not very good anticipating things beyond the curve.
Without the packaged best practices of the day, on-going improvements, a proven, scalable architecture, and with longer gestation periods and larger upfront costs, custom solutions will find it impossible to compete against the packaged ones. I am not saying packaged solutions do not have deficiencies – they certainly do. But a successful product gets created by a lot of experience in the market that (hopefully) translates into better flexibility and scalability, addresses things you are yet to think about and have a better prediction of the future.
If I can twist and stretch this comparison – we are seeing the same argument repeated in comparing the older Siebel CRM packaged solution with more contemporary solutions. Siebel is “complex”, “needs more upfront costs”, and is getting beaten by “simpler”, “easy-to-implement solutions” of today.
What do you think? Is there still an argument?
ps: btw. I have switched to a new Quora id nowadays.