4 Prompt Tracking Mistakes — Whiteboard Friday

Speaking of which, I deliberation the 2nd large correction that group make is reasoning astir this successful accepted SEO metrics, for illustration ranking.

So here, I could say, "Oh, fine, I'm mentioned first. Therefore, that's great. I've won." But actually, the Samsung brands person been mentioned much times.

Over a ample number of prompts, I mightiness beryllium acold much willing successful really galore times I'm mentioned astatine all, what percent of responses mentioned maine astatine all, really I'm mentioned, what are the snippets successful which I'm mentioned, and what benignant of adjectives are used.

There are each sorts of metrics. I deliberation this is rather a young benignant of abstraction correct now. So this is emerging successful position of what metrics group prefer. But I really deliberation you shouldn't attraction connected the accepted SEO ones because they're not that adjuvant successful this space.

Sure, it is simply a portion of contextual accusation who is mentioned first. It is thing you mightiness wish to look at, but it shouldn't beryllium your be-all and end-all successful the aforesaid measurement that it mightiness person been traditionally.

I would beryllium looking complete a ample number of prompts arsenic my benignant of breadstuff and butter, my balanced to ranking, earlier I get into the much subtle stuff. I would beryllium looking complete a ample number of prompts and conscionable looking astatine what percent of the responses mention maine astatine each arsenic my main benignant of gauge to get started, which brings maine connected to my 3rd large mistake, and that is quantity.

If you are a marque for illustration Apple, for illustration I've conscionable been talking about, you astir apt person been tracking. I presume that Apple does person an progressive SEO program, and they astir apt are search thousands, if not tens of thousands, possibly hundreds of thousands of keywords.

So it wouldn't make consciousness for a marque of that size to spell and say, "Okay, crossed each of our products and each of our brands and each of our markets, we're going to way 50 prompts."

If anything, location is much variance successful prompts. The outputs alteration much from time to day. They are little deterministic. They are longer tail. If anything, you would beryllium search much prompts than you were doing top-of-funnel keywords.

Now, don't get maine wrong. I deliberation if you're a mini business, I deliberation if you're a section business, I deliberation if you're a business that does a very circumstantial thing, possibly 50 prompts tin beryllium capable successful immoderate cases. But this is the aforesaid benignant of business that would person been search a smaller number of keywords arsenic well.

So I deliberation you request to person a standard that is appropriate, and that's what we're trying to support successful our products arsenic well, making judge that customers tin person a standard that is akin to their keyword search without it being benignant of astronomically costly aliases thing for illustration this. 

The past point I want to speak astir is caput terms. So erstwhile you're deciding what prompts to track, arsenic I mentioned, they're going to beryllium longer tail typically. I don't want you to return immoderate your existing keyword database was and conscionable dump it into immoderate your punctual search instrumentality of prime mightiness be.

Very fewer group are opening ChatGPT, typing "laptops," and past hitting Enter. I'm judge it's been done. It's astir apt been done by maine for an illustration connected a slide. But this is simply a weird point to do. This isn't really really group interact pinch these tools. So you should make judge you bespeak that successful the benignant of prompts that you're researching and including.

Now, punctual volume, it tin beryllium very difficult correct now. There are long-standing stats that group for illustration to use, which I deliberation is thing for illustration 15% of searches of keywords that Google sees each time are marque new, arsenic successful ne'er seen before.

One has to ideate that for prompts, this is 90-something percent. I don't cognize the nonstop stat. 

But if I were to guess, it's astir apt 90-something percent. I deliberation the overwhelming mostly of prompts astir apt person a measurement of astir one. So, really, you conscionable request to place the topics that are important and past person immoderate realistic prompts. 

We person immoderate devices to thief pinch this. My colleague, Dr. Pete, has written a great guide, which will beryllium referenced below, I hope. People Also Ask could besides beryllium a bully starting constituent because those are benignant of existent long-tail keywords that Google seems to deliberation beryllium and mightiness beryllium useful. There are tons of different things you tin do present to effort and get immoderate benignant of representative, much specific, longer-tail terms. Those are the 4 sorts of mistakes I wanted to talk about.

1. Consider specifying locations aliases personas

So 1 of the large ones that you could play pinch is locations and personas. 

So, successful your accepted rank tracking, thing that a batch of group do and thing that I've done successful the past is return the aforesaid keyword and past possibly way it from opportunity 50 different postcodes, aliases thing for illustration that, to spot if I get different responses. 

For different kinds of business, this whitethorn beryllium much aliases little useful. Or if you're not a business that deals successful real-world locations, past possibly this would beryllium much astir personas. So alternatively of saying, "What is the champion telephone for me?" you mightiness say, "What is the champion telephone for a assemblage student who has to do archive editing connected the go?" aliases thing for illustration this. That's a beautiful unusual example, but for example. Then you mightiness person various different personas, aliases you mightiness person various different locations, depending connected the type of business. 

2. Play astir pinch different languages

The 2nd point you tin do is play astir pinch languages. 

So astatine the moment, astir prompt-tracking devices are somewhat language-ambivalent because astir of the interfaces that they woody pinch are besides language-ambivalent. 

ChatGPT fundamentally responds successful immoderate connection you use. You tin unfastened ChatGPT. The interface is successful English. If you type to it successful Italian, it will reply to you successful Italian. So you tin return advantage of that. If you do beryllium successful aggregate markets wherever you want to spot what the responses are for illustration successful different languages, you tin return advantage of that and conscionable way prompts successful immoderate languages you spot fit. 

Obviously, that could beryllium mixed pinch the location point. 

Then these past 2 ideas are somewhat different. So this is little astir benignant of punctual description successful position of getting the punctual you want to way to get much item and to beryllium much representative, and much astir getting benignant of different types of insights alternatively than conscionable performance.

3. Specify qualities

So, for example, alternatively of saying what is the champion iPhone for blah, blah, blah, you could opportunity what is the cheapest phone, aliases the astir durable phone, aliases the astir trustworthy phone, aliases the astir privacy-respecting phone. You could person these arsenic benignant of different campaigns that you way successful bid to get different types of insights. 

4. Identify and attraction connected marketplace gaps

Or the flip broadside of that, this would return a small spot much processing connected your portion to spell done these insights, but you could ask, "What is it that the Samsung telephone does amended than the iPhone," aliases thing for illustration this, and past get benignant of much qualitative insights retired that you could past way complete clip to spot whether you're making a quality successful your trading pinch that perception. 

All right. I dream you for illustration these ideas. I dream you're not making these mistakes. If you are, I dream this was useful. Thank you very much. See you successful the adjacent Whiteboard Friday.