Chris Green
@chrisgreenseo
Followers
8K
Following
17K
Media
2K
Statuses
27K
Senior SEO Consultant (@TorquePartners), experienced search trainer & mentor. #BeOnePercentBetter
Colchester, UK
Joined February 2009
My "It Depends..." series hit 50 videos recently. I wanted to create content that is evergreen, pushed me out of my comfort zone & would be useful to other SEOs. I am looking for more questions - if I feature them in a future video you'll get a 3D printed speech bubble!
4
0
6
We're just a week away from the Free #SEOFOMO London Meetup 👀 Sponsored by @WixStudio & co-hosted by the awesome team of @StudioHawk 👇 * The event has been sold out since a while back: if you're registered, you should receive a reminder today! * We will have 3 insightful
2
14
30
I've launched this podcast to make SEO patents easier to understand! Every Friday, there's an episode on a patent I've picked and a 10-minute overview provided by NotebookLM. I've found it a nice way to absorb some info about patents on the move. https://t.co/ZmKMyzIMRJ
open.spotify.com
Podcast · Mark Williams-Cook · A weekly short discussion for SEOs that examines specific Google patents with discussions led by NotebookLM to make search-relevant patents easier to understand.
4
9
46
Oh yeah @RyanJones knows what's hot! I've just shipped one commercial CTR prediction model for a client. It takes in SERP snippet and position as input and outputs CTR delta. The model can tell which of many snippet variants will have the best click probability✨
This example isn't search, but this (click probability prediction) is what I think NavBoost is actually doing.
4
3
29
One of the easiest way to tell an SEO specialist experience is the way they tackle issues: A new SEO will try to fix all issues for a “perfectly optimized site”, while an experienced SEO knows that resources are limited, there’s limited flexibility and so many constraints, as
7
18
71
Should I do what an SEO blog post tells me? Here’s why I don’t think you should.
0
2
2
Some people are inclined NEVER to believe Google; I don't think that's probably all that helpful in the long run. The issue is, for those people who don't want to believe or trust it, can you blame them? Will it likely change significantly soon? Probably not. 🧵 4/4
0
0
1
BUT, the issue here is down to belief and whether you want to take what has been said as truthful or not. Google - from a Search PoV - has taken a lot of knocks this year and has been seen as misdirecting, misleading, or flat-out lying about certain elements. 🧵 3/4
1
0
0
I do believe that Danny/search engineers at Google probably do want to take action on the current state of SERPs. That and there are no specific up/down weighting factors just because someone is a "big brand"... 🧵 2/4
1
0
0
Props to Aleyda & Danny for this lengthy, thought-provoking conversation. They cover some challenging areas, particularly the status of brands following the HCU (helpful content updates), which is the most interesting and talked-about right now https://t.co/YHcp6NQuPN 🧵 1/4
1
2
3
Once you get used to writing prompts you'll find you don't need to break it down into so many stages - it will speed up. But I'd rather get the right answer slower, than getting the wrong output really quickly, which is the usual fear in using LLM outputs. 🧵 6/6
0
0
0
5️⃣ Review the detail to understand what's being done, challenge any issues with examples AND detail HOW it should be. There are more specific guides ( https://t.co/nzp60vFPfV), but this will help the process be less frustrating and increase your output's accuracy. 🧵 5/6
platform.openai.com
Explore developer resources, tutorials, API docs, and dynamic examples to get the most out of OpenAI's platform.
1
0
0
3️⃣ Give specific instructions around each point of the process, test each time, tweaking and getting it right. 4️⃣ If something is more complex or you don't know how to simplify any more, ask for a chain-of-thought response to see a "thinking out loud" style response. 🧵 4/6
1
0
0
Why not practice better prompt writing? 1️⃣ Start with a broad problem/question & ask for more info about the subject. If you're not sure where to start, this will give context. 2️⃣ Ask for a process to solve the problem or work to break it into different stages yourself. 🧵 3/6
1
0
0
Well, I'm not a fan of copy-and-paste prompts being offered up as THE solution ❌ Whilst you can get some mileage out of duplicating well-established prompts, the models behind them change and develop quickly, and SOME services handle some prompts better than others. 🧵 2/6
1
0
0
Prompt engineering - either a meaningless or frustrating term - is essentially a way to write more efficient prompts for an LLM like ChatGPT or similar. I've been asked lots recently about prompt guides/engineering; is there an essential list of prompts we should use etc? 🧵 1/6
1
0
1
A fairer marketplace could mean greater competition and innovation, so who knows what else may spring up to add additional value to the searcher or the search marketer. But then again, maybe Google appeals makes some concessions and avoids most of the fall-out 🤷♂️ 4/4
0
0
0
For the marketer who does most of their work through Google services, this may frustrate much of what you do and make things less effective/efficient. For the end user? At worst, you may see less relevant ads & parts of your search journey may feel clunkier (if noticeable). 3/4
1
0
0
Well… given the rise of AI in search results, the explosion of LLMs and potential new search competitors, the fragmentation of the search landscape and the deep ending role of social, I think the future of search has never been harder to anticipate. EVEN before this ruling. 2/4
1
0
0
So, it seems the day of reckoning for Google may have finally arrived…maybe. The DoJ wanting to break up the different elements of Google will be uncontroversial to most people, but what does this actually mean for the future of search? https://t.co/hraHQFYo5I 🧵 1/4
1
0
2
Will two businesses targeting the same geographical location within the same company compete with one another? You don’t have to answer with ‘it depends’, because the answer is broadly yes and no 🤣 Watch the video to see why.
0
0
0