SEO
How to Handle Out-of-Stock Products

Though it’s a joke in the SEO community, “it depends” is really the answer when dealing with out-of-stock products on e‑commerce websites. There are many situational decisions, with each having trade-offs, when it comes to SEO and user experience.
Many content management systems have built-in rules for handling out-of-stock products, but that logic can often be changed. Some setups may have no built-in logic at all. In these cases, you get to make the rules and work with your developers to implement what you want.
Let’s look at the different scenarios.
Permanently out of stock
With permanently out-of-stock products, you know the products are gone and won’t be back. Does that mean you should get rid of the pages? Not necessarily. You have a few options here.
Option 1. Redirect the page
If you have a similar product that you want to push people toward, you may want to 301 redirect the old product to the new product. This will also maintain any link value if the pages are similar enough.
If the pages are not similar enough or you redirect to a category page or your homepage, then these pages may be treated as a soft 404 and the value from links may not be preserved. If these other locations are where you want users to go, then do the redirect anyway. It’s possible the links will still pass value.
Because most websites have removed many products in the past, you may want to see if there are any opportunities to redirect old pages to relevant, current pages and reclaim the value from the links.
Here’s how to find those opportunities:
- Paste your domain into Ahrefs’ Site Explorer (also accessible for free in Ahrefs Webmaster Tools)
- Go to the Best by links report
- Use the response code filter and check “404 not found” and “410 gone”
I usually sort this by “Referring domains” or “UR” to look for top opportunities.

Best by links report.
When redirecting a page, many systems will automatically remove interna länkar from categories, facets, sitemaps, and internal search pages. In some systems, you may need to remove these links manually or there may be additional links from blog pages or elsewhere that you’ll want to remove.
To find these links, check the Links report in Ahrefs’ Site Audit, which is also free if you sign up for Ahrefs Webmaster Tools. Click into the area marked “Redirect” on internal links to get a list of the redirected links and the pages that have those links.

Internal links that redirect.
Option 2. Delete the page
If you are removing a page and there is no relevant, current product, you may want to simply delete the page and return a 404 or 410 status code.
As we mentioned in the first option, you also want to make sure that internal links to these pages are removed. This removal often happens automatically for many of the links to the page (but probably not all of them), so check for any remaining ones by crawling your site.
To find these links, check the Links report in Ahrefs’ Site Audit. Look for “Broken” on internal links. You can click into this chart to get a list of the broken links and the pages that link to them.

Broken internal links.
Option 3. Leave the page live
The product is gone, and most people will want to get rid of the page either by deleting it or redirecting it. But there are plenty of valid reasons to leave a page live.
It could have useful resources, such as documentation, that may take some burden off a support team. Perhaps, the page is still getting a decent amount of search traffic that you want to funnel into other products you offer. You can use Ahrefs’ Site Explorer to check the organic search traffic for the page.

If you’re leaving the page live for now, I don’t recommend removing internal links like you would in the other options. Removing the links can hurt your rankings.
This is also when potential SEO benefits clash with user experience. Landing on a page—where there’s nothing to buy—either from the internal search or from a category page won’t be a great experience for a user.
You may want to add a filter so users can remove out-of-stock products. Or you may want to show these products last in any listings. Deprioritizing the pages like this means they may also not rank as well, but I generally recommend it.
Eventually, you may consider these pages to no longer be useful and, accordingly, may want to redirect or delete them.
Temporarily out of stock, coming back soon
You want to leave these pages live. If so, there are some useful actions you can take or automate in your system.
There are features that may be useful to you and your customers, such as estimates of when a product will be back in stock, a wait list, or a way to sign up to be notified when a product has been restocked. Some stores are even offering their customers lotteries to purchase items when they’re back in stock, e.g., the Newegg shuffle program.
Temporarily out of stock, may not be coming back
Sometimes, you have a product that you just don’t know when or if you’ll be carrying again. In all of these cases, I’d make the product less visible on the website so users are less likely to see them. Let’s look at the options.
Option 1. Leave it live
As I mentioned before, there are several reasons you may want to leave a page live, including capturing search traffic to funnel people to other products or using the page for support or documentation purposes.
Don’t delete the internal links in this situation. Eventually, you may want to redirect or delete these pages when you realize the relevant products are not going to come back or the pages are no longer useful.
Option 2. Noindex
I’ll say that using noindex is not my favorite option, and I don’t typically recommend it. It’s usually not the best idea to do this because it can cut off the flow of PageRank.
I only mention it because, in some systems, noindex is used as the trigger to stop the product from being shown to users. On the bright side, products marked noindex will come back in search results pretty quickly after Google recrawls the page.
Option 3. Leave it live for a while and delete or redirect it later
At some point, you may just want to make a decision to treat this product as if it’s not coming back and delete or redirect it. When you do this is up to you, but a lot of people use logic based on the demand or when a certain amount of time has passed. In the short term, I’d keep the internal links live—but you’ll want to clean them up later.
How to find out-of-stock products
You should have a list of out-of-stock products from some data source that handles your product inventory or possibly within the content management system’s backend. In case you don’t have that list, you may want to crawl the website with Ahrefs’ Site Audit and search the HTML code to find pages that have your out-of-stock message.

Search for out-of-stock products.
Slutgiltiga tankar
As you’ve seen, you have a lot of options when it comes to out-of-stock products. What I generally recommend is you set some rules that you’re comfortable with and just go with them. The logic you use is really up to you. Ultimately, there’s no perfect solution.
Message me på Twitter if you have any questions.
SEO
Optimera din SEO-strategi för maximal ROI med dessa 5 tips

Wondering what improvements can you make to boost organic search results and increase ROI?
If you want to be successful in SEO, even after large Google algorithm updates, be sure to:
- Keep the SEO fundamentals at the forefront of your strategy.
- Prioritize your SEO efforts for the most rewarding outcomes.
- Focus on uncovering and prioritizing commercial opportunities if you’re in ecommerce.
- Dive into seasonal trends and how to plan for them.
- Get tip 5 and all of the step-by-step how-tos by joining our upcoming webinar.
We’ll share five actionable ways you can discover the most impactful opportunities for your business and achieve maximum ROI.
You’ll learn how to:
- Identify seasonal trends and plan for them.
- Report on and optimize your online share of voice.
- Maximize SERP feature opportunities, most notably Popular Products.
Join Jon Earnshaw, Chief Product Evangelist and Co-Founder of Pi Datametrics, and Sophie Moule, Head of Product and Marketing at Pi Datametrics, as they walk you through ways to drastically improve the ROI of your SEO strategy.
In this live session, we’ll uncover innovative ways you can step up your search strategy and outperform your competitors.
Ready to start maximizing your results and growing your business?
Anmäl dig nu and get the actionable insights you need for SEO success.
Can’t attend the live webinar? We’ve got you covered. Register anyway and you’ll get access to a recording, after the event.
SEO
TikTok’s US Future Uncertain: CEO Faces Congress

During a five-hour congressional hearing, TikTok CEO Shou Zi Chew faced intense scrutiny from U.S. lawmakers about the social media platform’s connections to its Chinese parent company, ByteDance.
Legislators from both sides demanded clear answers on whether TikTok spies on Americans for China.
The U.S. government has been pushing for the divestiture of TikTok and has even threatened to ban the app in the United States.
Chew found himself in a difficult position, attempting to portray TikTok as an independent company not influenced by China.
However, lawmakers remained skeptical, citing China’s opposition to the sale of TikTok as evidence of the country’s influence over the company.
The hearing was marked by a rare display of bipartisan unity, with the tone harsher than in previous congressional hearings featuring American social media executives.
The Future of TikTok In The US
With the U.S. and China at odds over TikTok’s sale, the app faces two possible outcomes in the United States.
Either TikTok gets banned, or it revisits negotiations for a technical fix to data security concerns.
Lindsay Gorman, head of technology and geopolitics at the German Marshall Fund, said, “The future of TikTok in the U.S. is definitely dimmer and more uncertain today than it was yesterday.”
TikTok has proposed measures to protect U.S. user data, but no security agreement has been reached.
Addressing Concerns About Societal Impact
Lawmakers at the hearing raised concerns about TikTok’s impact on young Americans, accusing the platform of invading privacy and harming mental health.
Enligt Pew Research Center, the app is used by 67% of U.S. teenagers.
Critics argue that the app is too addictive and its algorithm can expose teens to dangerous or lethal situations.
Chew pointed to new screen time limits and content guidelines to address these concerns, but lawmakers remained unconvinced.
Sammanfattningsvis
The House Energy and Commerce Committee’s hearing on TikTok addressed concerns common to all social media platforms, like spreading harmful content and collecting massive user data.
Most committee members were critical of TikTok, but many avoided the typical grandstanding seen in high-profile hearings.
The hearing aimed to make a case for regulating social media and protecting children rather than focusing on the national security threat posed by the app’s connection to China.
If anything emerges from this hearing, it could be related to those regulations.
The hearing also allowed Congress to convince Americans that TikTok is a national security threat that warrants a ban.
This concern arises from the potential for the Chinese government to access the data of TikTok’s 150 million U.S. users or manipulate its recommendation algorithms to spread propaganda or disinformation.
However, limited public evidence supports these claims, making banning the app seem extreme and potentially unnecessary.
As events progress, staying informed is crucial as the outcome could impact the digital marketing landscape.
Featured Image: Rokas Tenys/Shutterstock
Full replay of congressional hearing available on Youtube.
SEO
Allt du behöver veta

Google has just released Bard, its answer to ChatGPT, and users are getting to know it to see how it compares to OpenAI’s artificial intelligence-powered chatbot.
The name ‘Bard’ is purely marketing-driven, as there are no algorithms named Bard, but we do know that the chatbot is powered by LaMDA.
Here is everything we know about Bard so far and some interesting research that may offer an idea of the kind of algorithms that may power Bard.
What Is Google Bard?
Bard is an experimental Google chatbot that is powered by the LaMDA large language model.
It’s a generative AI that accepts prompts and performs text-based tasks like providing answers and summaries and creating various forms of content.
Bard also assists in exploring topics by summarizing information found on the internet and providing links for exploring websites with more information.
Why Did Google Release Bard?
Google released Bard after the wildly successful launch of OpenAI’s ChatGPT, which created the perception that Google was falling behind technologically.
ChatGPT was perceived as a revolutionary technology with the potential to disrupt the search industry and shift the balance of power away from Google search and the lucrative search advertising business.
On December 21, 2022, three weeks after the launch of ChatGPT, the New York Times reported that Google had declared a “code red” to quickly define its response to the threat posed to its business model.
Forty-seven days after the code red strategy adjustment, Google announced the launch of Bard on February 6, 2023.
What Was The Issue With Google Bard?
The announcement of Bard was a stunning failure because the demo that was meant to showcase Google’s chatbot AI contained a factual error.
The inaccuracy of Google’s AI turned what was meant to be a triumphant return to form into a humbling pie in the face.
Google’s shares subsequently lost a hundred billion dollars in market value in a single day, reflecting a loss of confidence in Google’s ability to navigate the looming era of AI.
How Does Google Bard Work?
Bard is powered by a “lightweight” version of LaMDA.
LaMDA is a large language model that is trained on datasets consisting of public dialogue and web data.
There are two important factors related to the training described in the associated research paper, which you can download as a PDF here: LaMDA: Language Models for Dialog Applications (read the abstract here).
- A. Safety: The model achieves a level of safety by tuning it with data that was annotated by crowd workers.
- B. Groundedness: LaMDA grounds itself factually with external knowledge sources (through information retrieval, which is search).
The LaMDA research paper states:
“…factual grounding, involves enabling the model to consult external knowledge sources, such as an information retrieval system, a language translator, and a calculator.
We quantify factuality using a groundedness metric, and we find that our approach enables the model to generate responses grounded in known sources, rather than responses that merely sound plausible.”
Google used three metrics to evaluate the LaMDA outputs:
- Sensibleness: A measurement of whether an answer makes sense or not.
- Specificity: Measures if the answer is the opposite of generic/vague or contextually specific.
- Interestingness: This metric measures if LaMDA’s answers are insightful or inspire curiosity.
All three metrics were judged by crowdsourced raters, and that data was fed back into the machine to keep improving it.
The LaMDA research paper concludes by stating that crowdsourced reviews and the system’s ability to fact-check with a search engine were useful techniques.
Google’s researchers wrote:
“We find that crowd-annotated data is an effective tool for driving significant additional gains.
We also find that calling external APIs (such as an information retrieval system) offers a path towards significantly improving groundedness, which we define as the extent to which a generated response contains claims that can be referenced and checked against a known source.”
How Is Google Planning To Use Bard In Search?
The future of Bard is currently envisioned as a feature in search.
Google’s announcement in February was insufficiently specific on how Bard would be implemented.
The key details were buried in a single paragraph close to the end of the blog announcement of Bard, where it was described as an AI feature in search.
That lack of clarity fueled the perception that Bard would be integrated into search, which was never the case.
Google’s February 2023 announcement of Bard states that Google will at some point integrate AI features into search:
“Soon, you’ll see AI-powered features in Search that distill complex information and multiple perspectives into easy-to-digest formats, so you can quickly understand the big picture and learn more from the web: whether that’s seeking out additional perspectives, like blogs from people who play both piano and guitar, or going deeper on a related topic, like steps to get started as a beginner.
These new AI features will begin rolling out on Google Search soon.”
It’s clear that Bard is not search. Rather, it is intended to be a feature in search and not a replacement for search.
What Is A Search Feature?
A feature is something like Google’s Knowledge Panel, which provides knowledge information about notable people, places, and things.
Google’s “How Search Works” webpage about features explains:
“Google’s search features ensure that you get the right information at the right time in the format that’s most useful to your query.
Sometimes it’s a webpage, and sometimes it’s real-world information like a map or inventory at a local store.”
In an internal meeting at Google (reported by CNBC), employees questioned the use of Bard in search.
One employee pointed out that large language models like ChatGPT and Bard are not fact-based sources of information.
The Google employee asked:
“Why do we think the big first application should be search, which at its heart is about finding true information?”
Jack Krawczyk, the product lead for Google Bard, answered:
“I just want to be very clear: Bard is not search.”
At the same internal event, Google’s Vice President of Engineering for Search, Elizabeth Reid, reiterated that Bard is not search.
She said:
“Bard is really separate from search…”
What we can confidently conclude is that Bard is not a new iteration of Google search. It is a feature.
Bard Is An Interactive Method For Exploring Topics
Google’s announcement of Bard was fairly explicit that Bard is not search. This means that, while search surfaces links to answers, Bard helps users investigate knowledge.
The announcement explains:
“When people think of Google, they often think of turning to us for quick factual answers, like ‘how many keys does a piano have?’
But increasingly, people are turning to Google for deeper insights and understanding – like, ‘is the piano or guitar easier to learn, and how much practice does each need?’
Learning about a topic like this can take a lot of effort to figure out what you really need to know, and people often want to explore a diverse range of opinions or perspectives.”
It may be helpful to think of Bard as an interactive method for accessing knowledge about topics.
Bard Samples Web Information
The problem with large language models is that they mimic answers, which can lead to factual errors.
The researchers who created LaMDA state that approaches like increasing the size of the model can help it gain more factual information.
But they noted that this approach fails in areas where facts are constantly changing during the course of time, which researchers refer to as the “temporal generalization problem.”
Freshness in the sense of timely information cannot be trained with a static language model.
The solution that LaMDA pursued was to query information retrieval systems. An information retrieval system is a search engine, so LaMDA checks search results.
This feature from LaMDA appears to be a feature of Bard.
The Google Bard announcement explains:
“Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence, and creativity of our large language models.
It draws on information from the web to provide fresh, high-quality responses.”
LaMDA and (possibly by extension) Bard achieve this with what is called the toolset (TS).
The toolset is explained in the LaMDA researcher paper:
“We create a toolset (TS) that includes an information retrieval system, a calculator, and a translator.
TS takes a single string as input and outputs a list of one or more strings. Each tool in TS expects a string and returns a list of strings.
For example, the calculator takes “135+7721”, and outputs a list containing [“7856”]. Similarly, the translator can take “hello in French” and output [‘Bonjour’].
Finally, the information retrieval system can take ‘How old is Rafael Nadal?’, and output [‘Rafael Nadal / Age / 35’].
The information retrieval system is also capable of returning snippets of content from the open web, with their corresponding URLs.
The TS tries an input string on all of its tools, and produces a final output list of strings by concatenating the output lists from every tool in the following order: calculator, translator, and information retrieval system.
A tool will return an empty list of results if it can’t parse the input (e.g., the calculator cannot parse ‘How old is Rafael Nadal?’), and therefore does not contribute to the final output list.”
Here’s a Bard response with a snippet from the open web:

System för konversation av frågor och svar
Det finns inga forskningsartiklar som nämner namnet "Bard."
Det finns dock en hel del ny forskning relaterad till AI, inklusive av forskare associerade med LaMDA, som kan ha en inverkan på Bard.
Följande hävdar inte att Google använder dessa algoritmer. Vi kan inte med säkerhet säga att någon av dessa tekniker används i Bard.
Värdet av att veta om dessa forskningsdokument är att veta vad som är möjligt.
Följande är algoritmer som är relevanta för AI-baserade frågesvarssystem.
En av författarna till LaMDA arbetade med ett projekt som handlar om att skapa träningsdata för ett konversationsinformationshämtningssystem.
Du kan ladda ner forskningsrapporten 2022 som PDF här: Dialog Inpainting: Förvandla dokument till dialoger (och läs abstrakt här).
Problemet med att träna ett system som Bard är att fråge-och-svar-datauppsättningar (som datauppsättningar som består av frågor och svar som finns på Reddit) är begränsade till hur människor på Reddit beter sig.
Det omfattar inte hur människor utanför den miljön beter sig och vilka typer av frågor de skulle ställa, och vad de korrekta svaren på dessa frågor skulle vara.
Forskarna undersökte att skapa ett system som läste webbsidor och använde sedan en "dialog inpainter" för att förutsäga vilka frågor som skulle besvaras av en given passage inom det som maskinen läste.
Ett avsnitt på en pålitlig Wikipedia-webbsida som säger "Himlen är blå" skulle kunna förvandlas till frågan "Vilken färg har himlen?"
Forskarna skapade sin egen datauppsättning med frågor och svar med hjälp av Wikipedia och andra webbsidor. De kallade datamängderna WikiDialog och WebDialog.
- WikiDialog är en uppsättning frågor och svar som härrör från Wikipedia-data.
- WebDialog är en datauppsättning som härrör från webbsidedialogen på internet.
Dessa nya datauppsättningar är 1 000 gånger större än befintliga datauppsättningar. Vikten av det är att det ger samtalsspråksmodeller en möjlighet att lära sig mer.
Forskarna rapporterade att denna nya datamängd hjälpte till att förbättra konversationsfrågesvarssystem med över 40%.
Forskningsdokumentet beskriver framgången med detta tillvägagångssätt:
"Viktigt, vi finner att våra målade datauppsättningar är kraftfulla källor för träningsdata för ConvQA-system...
När de används för att förträna standardretriever- och rerankerarkitekturer, avancerar de toppmoderna över tre olika ConvQA-hämtningsmätvärden (QRECC, OR-QUAC, TREC-CAST), och ger upp till 40% relativa vinster på standardutvärderingsmått...
Anmärkningsvärt nog finner vi att bara förträning på WikiDialog möjliggör en stark noll-shot-hämtningsprestanda – upp till 95% av en finjusterad retrievers prestanda – utan att använda någon ConvQA-data inom domänen. "
Är det möjligt att Google Bard tränades med hjälp av WikiDialog- och WebDialog-datauppsättningarna?
Det är svårt att föreställa sig ett scenario där Google skulle vidarebefordra träning av en konversations-AI på en datauppsättning som är över 1 000 gånger större.
Men vi vet inte säkert eftersom Google inte ofta kommenterar dess underliggande teknologier i detalj, förutom vid sällsynta tillfällen som för Bard eller LaMDA.
Stora språkmodeller som länkar till källor
Google publicerade nyligen en intressant forskningsartikel om ett sätt att få stora språkmodeller att citera källorna för sin information. Den första versionen av tidningen publicerades i december 2022 och den andra versionen uppdaterades i februari 2023.
Denna teknik kallas experimentell från och med december 2022.
Du kan ladda ner PDF-filen av tidningen här: Tillskrivna frågesvar: Utvärdering och modellering för tillskrivna stora språkmodeller (Läs Google abstrakt här).
Forskningsdokumentet anger avsikten med tekniken:
"Stora språkmodeller (LLM) har visat imponerande resultat samtidigt som de kräver liten eller ingen direkt övervakning.
Vidare finns det allt fler bevis för att LLM kan ha potential i informationssökande scenarier.
Vi tror att förmågan hos en LLM att tillskriva texten som den genererar sannolikt kommer att vara avgörande i den här miljön.
Vi formulerar och studerar Attributed QA som ett viktigt första steg i utvecklingen av tillskrivna LLM.
Vi föreslår ett reproducerbart utvärderingsramverk för uppgiften och jämför en bred uppsättning arkitekturer.
Vi tar mänskliga anteckningar som en guldstandard och visar att en korrelerad automatisk metrik är lämplig för utveckling.
Vårt experimentella arbete ger konkreta svar på två nyckelfrågor (Hur mäter man attribution? och Hur bra presterar nuvarande toppmoderna metoder på attribution?), och ger några tips om hur man kan lösa en tredje (Hur man bygga LLM med attribution?).
Den här typen av stora språkmodeller kan träna ett system som kan svara med stödjande dokumentation som teoretiskt säkerställer att svaret är baserat på något.
Forskningsdokumentet förklarar:
"För att utforska dessa frågor, föreslår vi Attributed Question Answering (QA). I vår formulering är input till modellen/systemet en fråga, och utdata är ett (svar, attribution) par där svaret är en svarssträng och attribution är en pekare till en fast korpus, t.ex. av stycken.
Den returnerade tillskrivningen bör ge stödjande bevis för svaret."
Den här tekniken är speciellt avsedd för frågor som svarar på frågor.
Målet är att skapa bättre svar – något som Google förståeligt nog skulle vilja ha för Bard.
- Attribution tillåter användare och utvecklare att bedöma "tillförlitligheten och nyansen" av svaren.
- Attribution tillåter utvecklare att snabbt granska kvaliteten på svaren eftersom källorna tillhandahålls.
En intressant anteckning är en ny teknik som kallas AutoAIS som starkt korrelerar med mänskliga bedömare.
Med andra ord kan den här tekniken automatisera mänskliga bedömares arbete och skala processen för att betygsätta svaren som ges av en stor språkmodell (som Bard).
Forskarna delar:
"Vi anser att mänskligt betyg är guldstandarden för systemutvärdering, men finner att AutoAIS korrelerar bra med mänskligt omdöme på systemnivå, och erbjuder lovande som ett utvecklingsmått där mänskligt betyg är omöjligt, eller till och med som en bullrig träningssignal. "
Denna teknik är experimentell; den används förmodligen inte. Men det visar en av riktningarna som Google utforskar för att producera pålitliga svar.
Forskningsuppsats om redigering av svar för fakta
Slutligen finns det en anmärkningsvärd teknologi som utvecklats vid Cornell University (också från slutet av 2022) som utforskar ett annat sätt att källattribut för vad en stor språkmodell ger ut och till och med kan redigera ett svar för att rätta sig själv.
Cornell University (som Stanford University) licensierar teknik relaterade till sök och andra områden, tjänar miljontals dollar per år.
Det är bra att hänga med i universitetsforskningen eftersom den visar vad som är möjligt och vad som är spjutspets.
Du kan ladda ner en PDF av tidningen här: RARR: Undersöka och revidera vad språkmodeller säger, med hjälp av språkmodeller (och read the abstract here).
Sammanfattningen förklarar tekniken:
"Språkmodeller (LM) utmärker sig nu i många uppgifter som inlärning av få skott, svar på frågor, resonemang och dialog.
Men ibland genererar de innehåll som inte stöds eller vilseledande.
En användare kan inte enkelt avgöra om deras utdata är tillförlitliga eller inte, eftersom de flesta LM:er inte har någon inbyggd mekanism för att tillskriva externa bevis.
För att möjliggöra attribution samtidigt som alla kraftfulla fördelar med den senaste generationens modeller bevaras, föreslår vi RARR (Retrofit Attribution using Research and Revision), ett system som 1) automatiskt hittar attribution för utdata från vilken textgenereringsmodell som helst och 2) efterredigerar utdata för att fixa innehåll som inte stöds samtidigt som originalutdata bevaras så mycket som möjligt.
...vi finner att RARR förbättrar attributionen avsevärt samtidigt som den ursprungliga indata bevaras i mycket högre grad än tidigare utforskade redigeringsmodeller.
Dessutom kräver implementeringen av RARR bara en handfull träningsexempel, en stor språkmodell och standardwebbsökning.”
Hur får jag tillgång till Google Bard?
Google accepterar för närvarande nya användare för att testa Bard, som för närvarande är märkt som experimentell. Google lanserar åtkomst för Bard här.

Google är på posten och säger att Bard inte är sök, vilket borde lugna dem som känner oro inför AI:s gryning.
Vi befinner oss vid en vändpunkt som inte liknar någon annan vi sett på kanske ett decennium.
Att förstå Bard är till hjälp för alla som publicerar på webben eller utövar SEO eftersom det är bra att veta gränserna för vad som är möjligt och framtiden för vad som kan uppnås.
Fler resurser:
Utvald bild: Whyredphotographor/Shutterstock
-
WORDPRESS2 dagar ago
Intern länkning för SEO: Den ultimata guiden för bästa praxis
-
AMAZON4 dagar ago
De 10 bästa fördelarna med Amazon AWS Lightsail: Varför det är ett utmärkt val för företag
-
SÖKMOTORER7 dagar ago
Uppdatering av Google AdSense Auto Ads Publisher Console
-
WORDPRESS6 dagar ago
De bästa webbhotelllösningarna för din personliga webbsida eller företagswebbplats
-
WORDPRESS6 dagar ago
ActivityPub för WordPress går med i Automattic-familjen – WordPress.com News
-
SÖKMOTORER1 dag ago
Google Search Status Dashboard Lägger till Google Ranking Updates
-
PPC4 dagar ago
PPC Campaign Testing: Dos & Don'ts för att förvandla risker till belöningar
-
MARKNADSFÖRING5 dagar ago
Hur man beräknar kundens livstidsvärde och maximerar det för ditt företag