Hi Jill and welcome to WebmasterWorld!

Google has been serving personalised results for a quite some time. This means that the SERPs will depend on whether you are signed in to Google or not, what your past searches are, at what location you are based and others.

When you do search for your keywords, do you clear your cookies first? This could be one of reasons why you see a different result on browser you are normally using versus browser you do not use or use rarely.

There is also possibility that your queries could be hitting a different data centre which could result on different SERPs you are seeing.

And lastly, Google has been in the state of flux with constant algo tweaks and whilst some sites do not feel it, others have been affected by it much more.

But more importantly, what does your web site analytics say about your traffic? Do you see any significant drops in your traffic or shift of traffic from certain keywords to others?

Thank you for your reply. This is what I was thinking was happening, just wasn't sure. We do not get a lot of traffic but what we do get is very targeted. If I could double that little bit it would be awesome but since I don't know what I'm doing with this any more I would be wary of changing anything drastically on the site for fear we'd lose what we do have. I notice there's a LOT of DIY results coming back for my searches, but we sell product, not recipes or instructions.

Having highly targeted traffic is great. You could do a futher keyword research to find out what other (highly targeted) keywords may apply to your site and develop new pages targeting these new keywords, whilst leaving your current pages alone the way they are now.

In the past few months when searching keyword terms for my website/industry we have come up on the first page for many terms often only to be no where to be found two days later and then back on top 24 hours after that.

I can't remember the exact source right now, but there was a comment by Matt Cutts about results not clicked on being less likely to show to the same visitor on subsequent searches [or on deeper pages of the same search] or something along those lines, and those comments make me wonder if that mechanism could be the cause of some of the randomness some of us have seen, because I've seen it myself.

When the preceding is combined with:

When you do search for your keywords, do you clear your cookies first? This could be one of reasons why you see a different result on browser you are normally using versus browser you do not use or use rarely.

It makes quite a bit of sense to me to think there's something algorithmically [technically heuristically] changing the results based on the previous clicks of a user [or browser] so I think cookies, logged-in v. logged-out, and even browser sniffing could definitely have an impact on what's seen in the results by any given user [or browser].

There is also possibility that your queries could be hitting a different data centre which could result on different SERPs you are seeing.

This is also a definite possibility, in-my-opinion, but if it's been going on for months I'm not sure if "datacenter alignment issues" are necessarily the most likely cause, but at the same time, they certainly could be.

If you register your site with Google Webmaster Tools you can then get average ranking data, and you can filter by Google property or country.

It is NOT always accurate (sometimes it's wildly inaccurate) but it can be a useful source of data; you can at least compare it to observations you are able to make or have made for you.

I think that the ranking data in GWT reflects the current fluctuation of the serps. I am sure that personalization is apart of this fluctuation, but also data centers and maybe even search volume contribute to the instability of KW rankings.

The ranking data in WMT are weighed by all site's URLs that rank for the particular keyword if there was an impression for the keyword.

For example, let's say you check a keyword "blue widget" in WMT. If you have URL ranking at #1 and another URL ranking at #23, and there was an user that went to third page of SEERPs, so the URL at #23 got impression, then this URL on page 3 will contribute to calculation of average ranking.

The avg. position is weighted that way, but if you drill down to a specific keyword it will show which URLs rank for that search query as well as your impressions at which position along with clicks and CTR. I find this more useful than the avg. position and it was what I was referring to in my post above. I apologize for not making that clear.