Category: google

Posted on

Tinkering With WP-CLI, Google Cloud, and BlueHost

I’ve been spending the last few days wrapping the WP-CLI application – a command line program to automate the administration of WordPress installations – inside a Java app so I can automate some WordPress work. One of the major bottlenecks was fixing up the correct SSH string to connect to the various WordPress providers.

Initially I was having a bit of difficulty because I misread the wp-cli documentation and I thought the –user argument was the SSH username. When I got that fixed, it turns out that some WordPress hosting services, such as BlueHost, require you to contact their support to activate SSH. I had to connect my application to multiple WordPress hosting services, but in this post I’ll use BlueHost as an example since they’re fairly representative of the work I had to go through.

In BlueHost’s case, having support activate SSH support was a surprisingly painless process – it only took a quick 5 minute text chat where I verified my email address. To build out the proper SSH command, I also needed to look at details provided by cPanel:

Click to expand.

All the information you need to build the SSH string is in the General Information section in the above screenshot. Your SSH string should look like this:

php wp-cli.phar plugin list --ssh=<CURRENT_USER>@<SHARED_IP_ADDRESS>:2222/<HOME_DIRECTORY>/public_html --debug

The BlueHost account I’m using as an example is a shared WordPress account, so it listed a shared IP. Make sure to double check the port number (2222 in the above code sample) – the usual SSH port is 22, but BlueHost uses 2222 for shared accounts. Note that I’ve listed an additional folder under the user home directory; the home directory path only takes you to the user’s home directory, but wp-cli needs the path to the WordPress installation, which is usually under another folder (in this case /public_html/ ).

Posted on

Why You Need To Buy Google Ads – GrubHub Edition

An incredibly thought-provoking article showed up today, written by a current restaurant owner – but one that has experience in technology circles. I strongly recommend reading it: https://www.saddlebackbbq.com/how-google-doordash-grubhub-conspire-screw-local-restaurants .

Obviously the top-line theme is how GrubHub and DoorDash take a large proportion of the monies from online orders. But there is a lesser theme that I want to emphasize which is easy to lose in the outrage. Quoted from the article:

DoorDash pays Google an advertising fee to steal customers that are searching for our restaurant name “Saddleback BBQ” and they are redirecting them to their own page. From there a customer can purchase from any BBQ restaurant in Lansing…

From https://www.saddlebackbbq.com/how-google-doordash-grubhub-conspire-screw-local-restaurants

Here’s the takeaway: While SEO is important, it’s not the end of the line. You could be the top result for your keywords, for your own business name (which was the case for this BBQ restaurant), and competitors will still take business away from you by running ads on your own name. This shows the power of a Google ad, especially an ad within Google’s Knowledge Panel (the side panel that provides contextual information).

That’s why it’s so critical to buy Google Ads on even searches you’re the top result in: to get rid of competitor ads that are trying to redirect business that should be yours. It might sound like a lot of extra money to spend, but as long as you provide a interesting landing page for the user the additional cost should not be too much.

One of the recurring themes I like to touch on with this site is how important ads revenue is to Google. As Google optimizes their search to show more local and “hyperlocal” content, more space will open up to show ads. Businesses will have a choice: buy up those ad slots, or their competitors will.

That’s why you need to be buying Google ads.

Posted on

Acquisition Thoughts

Lately we’ve seen some aggressive moves by Microsoft to pick up developer mind-share. The purchase of GitHub was the opening salvo, but Microsoft made a number of smaller moves recently as well: private repositories available for free on GitHub (previously you needed a student or paid account to have private repos), the acquisition of NPM (Javascript package manager and registrar), and GitHub Actions: a way to automate developer workflows – similar to a developer focused version of IFTTT/Zapier. Part of the new GitHub Actions makes it easier and faster to deploy code on Microsoft Azure.

Recently GitHub announced a huge drop of new features, the most important of which is Codespaces (an online IDE) and Discussions (a place to host community discussions). It’s clear to everyone that Microsoft is playing the long game in its war against Amazon Web Services: Microsoft is buying up developer mindshare, making it easier and faster to discuss, manage, and deploy applications on Microsoft services.

This leaves Google’s cloud platform in a difficult bind: how to compete against all these offerings? AWS is by far the market leader in the cloud game, with Microsoft in a strong 2nd place position and having strong Enterprise-size deployment credentials, plus increasing ownership of the development process. We’re beginning to see Google’s counter moves: recently they announced the release of a code editor within Cloud Shell. Here’s a screenshot as of today:

Screenshot of Google Cloud Shell’s code editor. Admittedly it’s a bit boring – it’s not as fully featured as AWS’s Cloud9 online IDE.

The bottom line is, Google Cloud needs to start opening up its checkbook if they want to compete with the options Microsoft and Amazon are developing/purchasing. Here’s a couple of companies I think Google should seriously consider purchasing:

  1. GitLab – GitLab is the perfect counter to Microsoft’s purchase of GitHub; GitLab includes the social coding aspect of GitHub, plus automated integration tools to compete with GitHub Actions
  2. An online IDE – There are a number of online IDEs available. I’ve been trying out a number of them and personally, I quite like GitPod.
  3. AI Services – Google needs something to differentiate its cloud platform compared to Microsoft Azure and AWS. An interesting play would be to lead the burgeoning AI industry. There are a number of players in this space, but a good starting acquisition would be Diffbot – it supplies APIs for structured extraction of text from a webpage and understanding context.

Some other acquisitions that I think make sense, especially in light of the current economic troubles driving down valuations:

  1. A note service to enhance GSuite. Google Keep exists, but it’s not as advanced as Evernote.

    With a $1 billion valuation, Evernote may be too expensive to purchase, but there are a number of smaller competitors that still provide a great note-taking experience. A good example would be Notion.
  2. AirTable. I’m surprised that nobody has acquired AirTable yet – it’s a marvelous new take on how spreadsheets and databases can be visualized.
  3. Automation tooling. Services such as IFTTT, Zapier, Integromat, Automate.io are the “glue” that can connect disparate services together. For instance, you can configure a new WordPress post to be saved to Google Drive/Dropbox – or any of hundreds of different web services.

    A purchase of IFTTT or similar service immediately buys integration into many different web services, plus allows a deeper integration into Google products. Imagine making it easier to share your favorite YouTube clip to any social media you have.

Posted on

Googlebot Cannot Scroll; Infinite Scroll Doesn’t Help SEO

I saw a fascinating article in today’s Search Engine Journal: Google’s Martin Splitt Explains Why Infinite Scroll Causes SEO Problems. Read it for some background information, but the bottom line is that Googlebot (the Google web crawler/indexer) does not scroll web pages, which means that any content exposed via infinite scroll is not indexed.

Infinite scroll can also cause other problems: some screenshot browser addons and services have difficulties rendering infinite scroll web pages. It’s easy to love infinite scroll as a user since it gives the illusion of infinite content, but it can be a nightmare for automated services.

The fix: Make sure that all content on a web site can be accessed without using the infinite scroll function. Also, send Google a sitemap so it knows where all the valid URLs are: https://support.google.com/webmasters/answer/183668?hl=en .

Posted on

A Google Credit Card?

TechCrunch is covering an upcoming Google product, a debit card under the umbrella of Google Pay. `

Branded credit cards bring in a surprising amount of money – the Apple credit card is projected to earn Apple $1 billion annually with very little risk. I think it’s notable that this is a debit card and not a credit card – a debit card charges purchases against a checking account (in other words, money is there before you make your purchase) while credit card purchases are essentially a short term loan. By making it a debit card, Google reduces their risk by requiring that consumers already have money in their checking accounts to charge against. Additionally, Google and its partner bank earn interest on that checking account’s balance.

This is another way of diversifying Google’s income stream away from Ads; I wouldn’t be surprised if we see the Google credit link to benefits on other Google properties – for example, free or discounted YouTube premium, discounted Google Home devices, etc.

Posted on

RSS: YouTube Channel Feeds

I’ve remarked before about how I love Newsblur as a replacement for Google Reader. But Newsblur can also watch for new YouTube videos via YouTube RSS feeds!

RSS support is not always clearly advertised on YouTube, but it’s simple to access. In NewsBlur, right click a folder and select Add A Site To This Folder:

Newsblur screenshot: add a new site to this folder.

Then just insert the YouTube channel URL, and NewsBlur should load the newest YouTube videos!

Add a New Site option in NewsBlur

This is a quick and easy way for me to monitor a lot of YouTube channels at once.

Posted on

Time Magazine: Google Maps Memory Lane

Quite a few articles are making the rounds this weekend about people finding their deceased loved ones on Google Maps. Here’s one from Time Magazine: Viral Story Is Leading People Down the Sweetest Google Maps Memory Lane. Similar stories are available on CNN and Buzzfeed.

I love these types of articles – finding unexpected uses of technology to connect closer with family and friends. What I think really sells this application of Google Maps is that these are pictures of people in the middle of their everyday lives – they’re not posed, or idealized.

Posted on

Americans Rank A Google Internship Over A Harvard Degree

I saw this article from Forbes: Americans Rank A Google Internship Over A Harvard Degree and thought it was worth highlighting.

It’s important to do well in school, and having a high undergraduate GPA helps immensely if one decides to go on to graduate school, get a MBA, etc. But an internship or two on a resume helps dramatically in getting that first job – they count as experience, which means you can immediately apply for those jobs that ask for 1 – 3 years of experience. Internships also give you something to talk about during interviews – how you handled difficult situations, how you handle meetings, and all those personality-style questions.

IMO the biggest value of an internship is to understand the corporate environment: learn how to dress, how to interact with colleagues and superiors, and to be able to compare yourself talent-wise and see where you need improvement.

If you’re a Chicago area undergrad looking for an internship, I strongly recommend applying to CME Group and Chase – they pay and treat their interns very well.

Posted on

SEO Best Practices – Marking Missing Pages With 404

I saw this tweet today that deserved to be highlighted:

Tweet from Google Webmasters Twitter account: Return a 404 code for any pages that are removed.

In short: if you remove a web page, make sure your server is returning 404 to correctly indicate that the page is removed and the URL is invalid.

I see a lot of sites that – for invalid URLs – return a 200 status then an error message in the body of the response. That only serves to confuse crawlers (and there are more crawlers on the web than Google’s).

Posted on

Losing Visits & Revenue During A Domain Name Change

A fascinating article popped up yesterday which underlines the importance of SEO and domain names: From BetaKit – Looka lays off 80% of staff as failed rebrand from Logojoy cuts revenue in half. Read the article – it’s a concise story about how Logojoy rebranded and moved to a different domain name; due to a series of errors, the company lost 80% of its organic traffic. In other words: because of a name/domain change Google and other search engines lowered or removed LogoJoy from their search results.

The typical wisdom when moving domain names is that a site will lose 20-30% of visits coming from search engines for 5 – 7 months after the move. However LogoJoy made two major errors which helped to drastically decrease the visitors they saw:

  1. During the rebrand from LogoJoy to Looka, the company also added more services; initially they were only creating logos, but they also added additional services such as business card designs, social media support, etc. Adding services is great, but doing so simultaneously to a rename only serves to dilute the value of a site in the eyes of a search engine.
  2. The name LogoJoy quickly summarizes what the site is about: it’s a place to get logos and possibly other services related to branding. In short: it’s a great, easily-memorable name that also helps SEO since it includes the word “logo”. The new name “Looka” is ambiguous: you can easily imagine multiple different companies in many sectors having that name. In addition, “Looka” doesn’t help SEO: it’s not immediately connectable to branding, logo, social media, etc.

    It wouldn’t surprise me if at least 10-15% of the traffic loss was due to the unclear new name: losing “logo” from the site name and not replacing it with a similar word strongly linked with branding (for example: media, brand, public relations, etc).

These issues could have been fixed by multiple ways, some of them pricy, some of them not so much.

The most obvious solution is not to rebrand. LogoJoy could have kept their logo-generating business at logojoy.com, then spun up another site ( BrandJoy.com? MediaJoy? ) to host their additional sales of business cards, social media assistance, etc. Once both sites were established and running for at least 6 months, then they could have been merged under the Looka brand.

A pricy-but-possible solution for LogoJoy would be – considering they had millions of dollars from venture capital funding – to simply buy their way out of the problem. LogoJoy could have bought up Google, FB and Twitter advertising for keywords relating to branding: logo, social media, how to brand my site, etc. Although this would be a very expensive move: easily at least several million dollars if not more.

The easiest solution would be to keep the LogoJoy name and sell the additional services they wanted to offer (business cards, social media) under the LogoJoy name as well.

Bottom Line Summary: Be very careful when moving domain names. When moving domain names keep the focus on the domain change. Make sure to appropriately 301 (Moved Permanently) the old site to point to pages on the new site. But most importantly, keep the focus on the move. Don’t dilute the value of your site by trying to enter new areas. If you can afford it, buy ads on Google for keywords relating to your site – the ads can do double duty by (1) referring users to you new site and (2) informing users about the name change.