Instance Customization Level Comparison Across Organizations

[Konrad3]

Hello,

We’re looking to better understand how the level of ServiceNow platform customization, as indicated by the volume of records in the sys_update_xml table, compares across different organizations.

We recognize that this metric alone isn’t a definitive indicator of true customization level or general platform health, as it can be influenced by factors such as instance age, development practices, number of developers, and update set usage. That said, we’re hoping to gain some general insight from the community.

From your experience:

• How do different customization record count ranges (for example, ~100k, ~250k, ~500k, or 1M+) typically compare across organizations?

• Does ServiceNow provide any informal or internal benchmarking related to this metric (I couldn't find any personally)?

We’re not looking for precise comparisons, just high-level insight to help us better contextualize our customization footprint relative to other organizations.

Any perspectives or experiences you’re willing to share would be appreciated.

Thank you.

[Matthew_13 · Accepted solution]

Hi Buddy,

There isnt any official or even informal benchmarking from ServiceNow that says a certain record count is good or bad. In practice, most teams use this number only as context, not as a measure of platform health.

From what I generally see across customers, the ranges you mentioned tend to line up roughly like this:

• Around 100k is often a newer instance or one with tighter governance and lighter customization.

• Around 250k is very common for mid-maturity enterprise instances that have been in use for several years.

• Around 500k usually shows up in older or more heavily used environments with multiple teams and lots of historical changes.

• 1M+ is typically long-lived enterprise or MSP instances and almost always reflects years of accumulated work rather than “over-customization” by itself.

The important thing it isnt the absolute number, but things like:

• how fast that number is growing,

• what types of updates make it up UI, scripts, ACLs, BRs, etc.....

• how much of it is still relevant versus legacy noise,

• and whether customizations are creating upgrade pain or operational risk.

So Basically :
there’s no benchmark to compare against, and the raw count alone doesn’t tell you much. It’s useful for high-level context, but the real insight comes from trends, composition, and impact rather than the total number itself.

@Konrad3 - Please mark Accepted Solution and Thumbs Up if you found Helpful!

[Konrad3]

Thank you Matthew for your response. This is very helpful. I much appreciate that you provided insight on the different ranges of customizations.