Quote:
Originally Posted by Femjoy Michael
Agreed. I wished there was more business talk here too.
Alexa sample data is not accurate but does provide an over-view of trends. It takes into account ALL traffic, including garbage paid traffic.
If you know what you are benchmarking on google trends, you can get a pretty good estimate of "quality" traffic such as type-ins, directs, and customers willing to pay.
Let's say I know my competitor converts the same as I do. Since we are in the same niche, our traffic patterns will be similar for type-ins. I benchmark joymii.com against their URL. Knowing the exact amount of direct traffic I get, I can estimate how much direct traffic they get, and therefore roughly their sales amount, even their possible member count.
A paysite can have a really low Alexa, but high google trend since the data is normalized.
Let's say I was your competitor. One day I looked up autumnriley.com on alexa. I see nothing. But when I look it up on google trend, I see a huge spike upwards. That means you are getting a huge increase in direct and SE traffic - the good kind of traffic. I would then try to analyze other data such as SERPs, and WayBackMachine saves - anything to see what you have done to get that huge uptick and how I can get the same thing
|
But there is alot of search terms to get to your sites via google, not just one .. and you better hope that one search term isnt' used by people searching anything else up. And do you know your competitors top search terms? (or that weeks top search terms) You would need to know things like that too
I guess you are right about it taking into account garbage paid traffic but unless you are just spending a fortune nonstop for a long period, it hardly skews these trends and stats. sure it will show a spike and then that spike will go down, but like i said, when you go at the historical data, gives you a good estimation of what they usually sit at and what direction they are heading in. if you want to get more specific into the week, i think that is pointless other than if you want to see how popular an update or something was from them. then that would be similar to what i was saying about searching up popularity of a model (something more specific)
Its a shame we couldn't test our theories with a site by guessing the memberbase because i dont trust a single person to tell me the correct number so.. ;)
But I am much more confident in my ability to do so with alexa traffic stats and discounting any large spikes that could be caused by end of a promotion, etc
And using solo sites as an example is poor because alexa does not track stats of sites with that traffic level, which i stated in my first post.
Are you sure youre using reach percentage and not traffic rank stats, maybe thats why you feel its more accurate ;) Given you looked up autumn and you felt it was more accurate, i feel like you didn't use reach percentage, since there is no available data
If you are more confident in your ability with that method, by all means. But im not sure how well tested that is, i personally have ran 500+ tgps, combine for over a million a day in traffic. Checking alexa on them was all i did considering alot of tgps back in the day, you had to had over a certain rank, etc in order to trade.
Alexa was accurate on estimating all my sites to within 5k of traffic, and thats big. But thats because they were anywhere between 20k-100k day in traffic. With something bigger, it would be closer to 10-20 percent maybe but again, thats very accurate. From my experience, traffic = sales, most things in this industry convert pretty closely as the paysite business model is very very similar, even from niche to niche.
But i guess you are only thinking google traffic because i assume think thats the best way to determine their memberbase (which i dont agree with but yeah) then that just brings me back to my first paragraph again