4 ms·

This seems like a poor analysis to me. Did you control for other factors? It's likely that older domains have had more time to build links, trust, authority, e

by Roedou 14y ago

This seems like a poor analysis to me. Did you control for other factors? It's likely that older domains have had more time to build links, trust, authority, etc. It's entirely possible that Domain Authority has a more causal relationship with ranking position, and that domain age is unrelated.

dchuk 14y ago
We're not by any means claiming that we made a purely scientific experiment as you can't particularly run with controls against Google anyways. We're simply taking a look at the data our product has collected in the aggregate and publishing it and adding some interpretation of the data so that people can take what they wish from the data and apply it to their own campaigns.
- Roedou 14y ago
  It would be useful to compare to a raw graph of the domain age of your customers. There's a chance that it looks similar to this - and that even if Google picked the #1 ranking site at random, then this chart would look the same.
  - dchuk 14y ago
    sorry I don't think I understand what you mean by raw graph
_delirium 14y ago
I'm not even sure what the baseline is. What percentage of domains with any relevant content at all are >1 yr old? For it to be surprising that 93% of first results are from >1yo domains, it would have to be the case that >1yo domains make up much less than 93% of the relevant population, so we can conclude that they're actually overrepresented in the top results.
- dchuk 14y ago
  Good point. Unfortunately, we don't have any realistic means to collect that data so we can't really accommodate such an analysis. We're working with what our product has collected and doing our best to share that and apply our own SEO-based interpretation of the data. There are a handful of unsubstantiated claims that float through the SEO world that we're trying to provide some actual numbers for. If we could without a doubt show you what works in SEO we'd be sipping pina coladas on an island somewhere :). All we're saying with this series of posts is "out of all of the keywords people have analyzed using serpIQ, this is what we make of this specific assertion/data point/metric"

by Roedou 14y ago

dchuk 14y ago

We're not by any means claiming that we made a purely scientific experiment as you can't particularly run with controls against Google anyways. We're simply taking a look at the data our product has collected in the aggregate and publishing it and adding some interpretation of the data so that people can take what they wish from the data and apply it to their own campaigns.

Roedou 14y ago
It would be useful to compare to a raw graph of the domain age of your customers. There's a chance that it looks similar to this - and that even if Google picked the #1 ranking site at random, then this chart would look the same.
- dchuk 14y ago
  sorry I don't think I understand what you mean by raw graph

_delirium 14y ago

I'm not even sure what the baseline is. What percentage of domains with any relevant content at all are >1 yr old? For it to be surprising that 93% of first results are from >1yo domains, it would have to be the case that >1yo domains make up much less than 93% of the relevant population, so we can conclude that they're actually overrepresented in the top results.

dchuk 14y ago
Good point. Unfortunately, we don't have any realistic means to collect that data so we can't really accommodate such an analysis. We're working with what our product has collected and doing our best to share that and apply our own SEO-based interpretation of the data. There are a handful of unsubstantiated claims that float through the SEO world that we're trying to provide some actual numbers for. If we could without a doubt show you what works in SEO we'd be sipping pina coladas on an island somewhere :). All we're saying with this series of posts is "out of all of the keywords people have analyzed using serpIQ, this is what we make of this specific assertion/data point/metric"