Avg speed as gravel training metric?

This was inspired by listening to your podcast with Ian Boswell post unbound, in which he talks about focusing on improving his average speed during some of his training rides. I thought this was interesting as I’m more used to focusing on power and HR to #ftfp. I tried focusing on average speed for a couple rides between training cycles and found it to be a fun and intuitive way to do these rides.

What are your thoughts on using average speed to help guide training or incorporating it into a gravel plan? I am still a new racer, but the events I have done are 100ish mile, 6-10k’ of climbing gravel races. I am currently a “middle of the pack” finisher (though have improved a lot over the last year, thanks Gravel Training Plan!). In events like these, that means I spend a lot of time out there working alone. Curious on what your thoughts are about what kind of role average speed could play in something like the simulation rides or other training rides. Is it actually a trainable metric or is it more a reflection of improvement in the power-based training we do?


Average speed is a great training metric per the Ian Boswell podcast.

A few caveats - av speed is terrain dependent . The hillier your route the lower you av speed will be. Opposite for flat (higher at speeds)

I would encourage you to use / compare av sped only for the same ride routes. Wind and Groad conditions will also impact about speed.

All these caveats are why using power is still a good idea. Use both. And use power when the routes vary and this so will be av speeds too.

You can also create a Strava-Val and go by at speed for your chosen segment in addition to time.

Have fun -