Don McIntosh: With over 400 LMSes available, it seems foolish to me to build one from scratch. The process would take longer, be more expensive and have many headaches. If you have a large IT group with nothing to do, if the needs of your organization are truly unique, then it might make sense.
Donald H. Taylor: There are so many LMSes available it would be madness to write your own. Maintaining your LMS afterwards (dealing with scope creep, modifications and bug fixes) will be a full-time job. Spend the time you would have spent scoping an LMS on talking to organizations that are delighted with the learning technology they are using. You’ll soon find an LMS to fit.
Andy Wooler: In addition to the challenge of having to build and maintain it, the in-house system will need to integrate into the wider HR technology infrastructure such as talent, performance, HRIS and your HR data solution. For me, there’s a wider question here too, and that is why is your LMS broken/outdated in the first place? Did you customize it so much you couldn’t keep up with the updates? Have your business requirements changed since you implemented it? Is it part of a holistic solution for HR or does it sit in glorious isolation? Before going anywhere near an in-house build, revisit the business requirements and your current solution’s ability to be configured to meet new or changed needs.
Susan Fore: Years ago we were asked to cut operational costs and one way was to stop funding our LMS and having a replacement system created internally. The IT guy spent a couple of hours reviewing our current system and our requirements. He walked away shaking his head and said no way did they want to touch the project. We continued with the purchased system. Your money will be well spent to ensure you have the right requirements defined, then pick a system. There are a lot of suitable products. None are perfect!
Ironwood Learning: I used a custom LMS for 10 years and quickly discovered how difficult and expensive it is to keep the technology up to date and working for our customers. I would never recommend building or doing significant customizations to an LMS today. There are some very good products available that are very easy to use and configure for both administrative staff and the end user. Remember, is it all about delivering learning to the end user so we can measure the transfer of knowledge, improve skills, reduce errors and increase company profits. Keep it simple so learners can find, register and consume content to support their jobs.
Gary Wise: I’ve done both, built one from scratch and bought four others in previous lives. Don’t rush to build something. There is too much technology out there today to do that. With the way learning is converging with work, my last big purchase was an electronic performance support system that dramatically reduced the need for the LMS. You have options … and the technology is keeping pace with business needs. … It is we who are not keeping up because we are still dragging a big ol’ LMS around when Web 2.0 technology has the potential to change the rules of learning engagement in a very agile way at the point of work.
Peter Cridland: Another way to look at it is economies of scale. A reasonably successful vendor could have hundreds of customers or more. The vendor has invested significant in the flexibility and agility to meet the requirements of all those customers. Compare this with trying to build your own. You would need a large pool of freely available, cheap resources that can build and then support a production environment year after year. For any business that needs to maintain focus and competitive edge with its end customers, don’t waste your time and resources trying to replicate technology already available.
Gloria Pakravan: With so many vendors out there, the marketplace is confusing. But building your own would go from confusing to madness! The process that I like to use when choosing an LMS or replacing one is the following:
1. Document your current state.
2. Document your desired state.
3. Based on the above, put together a requirements list and determine what of the list is must haves vs. nice to haves.
4. Issue an RFP.
5. Have vendors demo the product.
6. Short list to top two to three vendors and ask for a sandbox, an environment the vendor provides for potential customers to play with the software. Using your desired state, create use cases and go through the sandbox to see if there is a fit.