OSS Vendor 1. “I have 1 million features.” (Dr Evil puts finger in mouth)
OSS Vendor 2. “Yeah, well I have 1,000,001 features in my OSS.”
This is the arms-race that we see in OSS, just like almost any other tech product. I imagine that vendors get into this arms-race because they wish to differentiate. Better to differentiate on functionality than price. If there’s a feature parity, then the only differentiator is price. We all know that doesn’t end well!
But I often ask myself a few related questions:
- Of those million features, how many are actually used regularly
- As a vendor do you have logging that actually allows you to know what features are being used
- Taking the Whale Curve perspective, even if being used, how many of those features are actually contributing to the objectives of the vendor
- Do they clearly contribute towards making sales
- Do customers delight in using them
- Would customers be irate if you removed them
Earlier this week, I spoke about a friend who created an alarm management tool by himself over a weekend. It didn’t have a million features, but it did have all of what I’d consider to be the most important ones. It did look like a lot of other alarm managers that are now on the market. The GUI based on alarm lists still pervades.
If they all look alike, and all have feature parity, how do you differentiate? If you try to add more features, is it safe to assume that those features will deliver diminishing returns?
But is an alarm list and the flicking of tickets the best way to manage network health?
What if, instead of seeking incremental improvement, someone went back to the most important requirements and considered whether the current approach is meeting those customer needs? I have a strong suspicion that customer feedback will indicate that there are definitely flaws to overcome, especially on high event volume networks.
Clever use of large data volumes provides a level of pre-cognition and automation that wasn’t available when simple alarm lists were first invented. This in turn potentially changes the way that operators can engage with network monitoring and management.
What if someone could identify a whole new user interface / approach that overcame the current flaws and exceeded the key requirements? Would that be more of a differentiator than adding a 1,000,002nd feature?
If you’re looking for a comparison, there were plenty of MP3 players on the market with a heap of features, many more than the iPod. We all know how that one played out!