3 ms·

Ask HN: Building your own CRM vs. Salesforce/SugarCRM

3 points by wuliwong 9y ago

We are going to start using SugarCRM for dealing with tracking buying/remodeling/selling of real estate. I am building a real estate AVM which will read and write data concerning valuations to this CRM. As I understand it, there will be a large amount of customization of the CRM required with a lot of custom modules built. At what point is it easier to just build your own web app from scratch rather than try to use something like SugarCRM? I would love to hear any personal experiences in either direction.

svennek 9y ago
Not really related to your domain, but... It is my experience that even if building on a standard platform is nice (and fast) in the beginning, the downsides show themselves fast: 1) If your world-view and that of the "standard, customizable system" (SCS) is somewhat different, you are going to fight it forever 2) SCS is most likely very hard to upgrade in the future (depends on lots of stuff, talking ballpark here). 3) Some stuff your customers ask for is going to be almost impossible to do (variant of 1). I have tried two projects where we tried to boost ahead by using a SCS-base (once wordpress+woocommerce and once drupal). Both failed spectacularly and ended up being scrapped due to eternal fighting against the grain of the SCS... notice, no blame is put on the SCSs themselves... The latter also ended up living long enough to also be a problem for updating... Regarding updating, with custom software (with custom bugs/security holes), upgrading can be easier, but it is also not as urgent as a zero-day on a common piece of OSS... Also if your SCS-upstream is going to deprecate the version you are building on, you are going to be in a world of hurt... I have been doing custom programming/integration for almost 20 years, so that is not my usual track-record (2 of 2 failed) My personal opinion, if you already know that your are going to heavily modify it, you would probably win by building the "right size" from scratch. (but I am, by heart, a builder and not a customizer) But in the end, it depends on a lot of stuff, including: - does your org have enough skill to build from scratch (which is much harder, as nobody has taken away the icky security problems) - is launching fast (i.e. run-way problems) or lowest cost in the long run the most valuable - what is the chance of doing a full rewrite in 3 months, if you find that SCS was the wrong approach And please, be extra careful about security and do it right... no shortcuts here... actually, start out by building it secure with no features, and add them later (vs. the opposite approach).
africajam 9y ago
I've built a real estate listings website from scratch using Ruby on Rails. It was harder than I expected but now I have a solid base from which I can build functionality quickly. The best part is that I open sourced it! Check it out and see if it can be useful for what you're working on: https://github.com/etewiah/property_web_builder https://github.com/etewiah/property_web_builder Happy to help with any questions you might have.

3 points by wuliwong 9y ago

svennek 9y ago

Not really related to your domain, but... It is my experience that even if building on a standard platform is nice (and fast) in the beginning, the downsides show themselves fast: 1) If your world-view and that of the "standard, customizable system" (SCS) is somewhat different, you are going to fight it forever 2) SCS is most likely very hard to upgrade in the future (depends on lots of stuff, talking ballpark here). 3) Some stuff your customers ask for is going to be almost impossible to do (variant of 1). I have tried two projects where we tried to boost ahead by using a SCS-base (once wordpress+woocommerce and once drupal). Both failed spectacularly and ended up being scrapped due to eternal fighting against the grain of the SCS... notice, no blame is put on the SCSs themselves... The latter also ended up living long enough to also be a problem for updating... Regarding updating, with custom software (with custom bugs/security holes), upgrading can be easier, but it is also not as urgent as a zero-day on a common piece of OSS... Also if your SCS-upstream is going to deprecate the version you are building on, you are going to be in a world of hurt... I have been doing custom programming/integration for almost 20 years, so that is not my usual track-record (2 of 2 failed) My personal opinion, if you already know that your are going to heavily modify it, you would probably win by building the "right size" from scratch. (but I am, by heart, a builder and not a customizer) But in the end, it depends on a lot of stuff, including: - does your org have enough skill to build from scratch (which is much harder, as nobody has taken away the icky security problems) - is launching fast (i.e. run-way problems) or lowest cost in the long run the most valuable - what is the chance of doing a full rewrite in 3 months, if you find that SCS was the wrong approach And please, be extra careful about security and do it right... no shortcuts here... actually, start out by building it secure with no features, and add them later (vs. the opposite approach).

africajam 9y ago

I've built a real estate listings website from scratch using Ruby on Rails. It was harder than I expected but now I have a solid base from which I can build functionality quickly. The best part is that I open sourced it! Check it out and see if it can be useful for what you're working on: https://github.com/etewiah/property_web_builder https://github.com/etewiah/property_web_builder Happy to help with any questions you might have.