Tracking words or sets

ok - Here's another question for yall.

Should I track my CTR and ROI on specific words or should I track sets of words?

Example:
track this: needles, knitting needles, crochet hooks, beading needles, etc. etc. etc.
or track this: needles
and this: knitting needles
and this: crochet hooks
and... etc. etc.

I had initialy set up my tracking to follow individual terms/phrases. but I found that it doesn't take too long for that to get way out of hand. Especially when you are bidding/advertising large amounts of words.

But now that I am turning quite a decent CTR and ROI - I was thinking I should start splitting them up from a big group and see where they are coming from - exactly.

Does anyone have suggestions on this? pros/cons on either option? I am considering upping my PCP dramatically, but before I do - I want to be sure I have things in place. Please offer me feedback that can help in my endevour. :]

Should I track my CTR and ROI on specific words or should I track sets of words?

I will have to go with specific words; deep-leveled quatifiable data are always helpful

Tracking the terms in a group will suffice a smaller campaign, but it appears to me that your campaign isn't all that small in nature.

>>> Hypothetical Situation 1 (HS1) <<<

Group 1="needles, knitting needles, beading needles"

Ad impression: 40,000
Clicks: 5,000

Cost of campaign: $3,000
Ad impression: 40,000
CTR: 0.125
Conversion: 0.5
Avg. Sales Per Person: $500
ROI: $415.67

>>> Hypothetical Situation 2 (HS2) <<<

Search Term 1 - needles
Ad impression: 10,000
Clicks: 2,000

Cost: $1,000
Ad impression: 10,000
CTR: 0.2
Conversion: 0.35
Avg. Sales Per Person: $450
ROI: $314.00

Search Term 2 - knitting needles
Ad impression: 15,000
Clicks: 1,500

Cost: $1,000
Ad impression: 15,000
CTR: 0.1
Conversion: 0.5
Avg. Sales Per Person: $500
ROI: $374.00

Search Term 3 - beading needles
Ad impression: 15,000
Clicks: 1,500

Cost: $1,000
Ad impression: 15,000
CTR: 0.1
Conversion: 0.4
Avg. Sales Per Person: $450
ROI: $269.00

-----------------
(Above figures are made up for the sake of conversation)

Using HS1 and HS2 figures...planning for next month with same budget of $3,000 (also allowing appropriate figures to remain constant for the sake of conversation)

HS1 doesn't provide much information; budget distribution will have to rely on the person's intuition/gut-instinct.

HS2 does provide more information that can be factored in decision-making process. Search term 2 (knitting needles) has higher conversion and ROI - one might decide to distribute more fund into this search term while sacrificing the fund for search term 3.

Various decisions can be made in HS2, but the important point is that the person will be making his/her decision based on more concrete data than HS1.

But...

Do consider that there are more involved variables in more realistic settings.

If you were to lack resources to track individual search terms and acquring the appropriate resources will cost a significant amount of fund, tracking grouped search terms will be sufficient and more 'sensible'

If tracking individual search terms were to take up much of your time, do consider the alternatives (with opportunity cost in mind). After all, your time is money, and you might be able to pull in more $$$ doing something else other than tracking/analyzing individual search terms.

HTH
(do apologize for the sloppy writing...I wrote this in a hurry)

Wow! Now that's some great information!

I could see the sense, when a campaign gets too large, to track by groups on a regular basis, and to take the time to dig into a group now and then to weed out the under-performers.

For any campaign of managable size, I'd certainly opt for tracking by the individual word or phrase, especially since some new programs (like ClickTracks) now make it so much easier to do this kind of tracking.

I could see the sense, when a campaign gets too large, to track by groups on a regular basis, and to take the time to dig into a group now and then to weed out the under-performers.

Excellent point on weeding out the under-performers from time to time, Jenn

Thanks pointing out that out!