Exploring Changing Conversion Rate Of Google Traffic

[goodroi]

There are many reports of Google traffic converting at different levels than previous experiences.

For the sites experiencing a change, is it because visitors are seeking something else or are they being incorrectly directed? Is Adwords siphoning certain types of users which changes the audience profile that eventually get to organic listings? Has voice search or other search avenues impacted online activity? Has Google stayed the same but the internet user base has evolved to behave differently? Or do you think these these reports are isolated cases?

[aristotle]

Well if a visitor doesn't convert,

1. perhaps the visitor was mis-matched to that site and from the beginning there was no chance at all to get a conversion.

2. the visitor might be a potential conversion, but deficiencies in the website caused the visitor to leave and go elsewhere.

[iamlost]

Such an innocent simple topic... rather like an iceberg, where 90% is out of site but responsible for the visibility of the noticed 10%...

Sooo, I shall take the time to shine a brief light on the complexity below the surface, largely ignored as unseen without analytics.

1. Bots:
Non human traffic makes up 20% to over 90% of a site's traffic volume; typically the smaller the site the greater the bot percentage.

Google Analytics and most hosts typically miss half of today's bots. For sites with low traffic numbers this mixture of bots can make less than monthly conversion numbers look inconsistent and/or lead to high claw backs of third party ad payments. Further, as these more stealthy bots proliferate such problems will be progressively exacerbated.

2. Mobile.
With the advent of mobile and especially as, typically, visitors are as or even more likely to come via mobile devices conversion/revenue takes a severe one two punch.

While it varies by niche mobile visitors are more likely (in comparison with desktop) to be in the research rather than buy phase.

This means that it is critical that content not simply be designed to sell convert but to (1) lead the visitor from the research through purchase process all in one visit or (2) be compelling enough to bring them back later to complete the sale (and you need be able to recognise the return).

Either is far different, in kind not just degree, than default SEOed content designed for one time first time SE referrals.

The second punch is that third party ad networks pay out less for mobile visitors. This means that mobile visitors, on average, are worth one to two thirds desktop.

3. Traffic quality
This is particularly iffy with dotcom as it has both a global generic reach and a US-centric commercial history. Thus if one has a dotcom domain yet are constrained to shipping only within the lower 48 US states traffic from Alaska or Canada, Argentina or Bulgaria is non-convertable.

When not using a cctld it is critical to maximise other signals to indicate boundaries/constraints while understanding that SEs will get it wrong at least some of the time.

Even with cctld's the more local a site's target audience the more similar to the foregoing tld problem.

In the old days one could query specific SE data centres to investigate differences in query resolution. While that sort of directed query is largely gone it does appear that the data centre queried has an unappreciated influence on the traffic referred.

The above and more need to be considered when looking at whether or how Google organic traffic conversion changes are occurring on any specific site.

Note: plus, of course, the continuing increasing retention on referral to Google properties, niche audience query change over time, etc.

In short, pretty much some of every question raised. However, the degree varies by niche and even within a given niche.

It would be safe to consider things to proceed as have been aka things will get worse vis-à-vis Google organic traffic.

[brighteryeg]

You have the option to filter bot traffic in Google Analytics but I haven't done much testing with it. I'm not sure how effective it actually is.