Feb 28, 2017
Feb 23, 2017
What do productivity, machine learning and next generation teams have in common? Google Cloud Next ‘17.
On March 8-10, Google will host one its largest events ever — Google Cloud Next 2017. In the last year, the Google Cloud team has introduced some new products and solutions to help businesses face some of their biggest productivity problems. Next is our way of bringing together customers and partners under one roof to see the results of all these updates. That includes the latest cloud innovations and more than 200 sessions, where you can check out new products and features firsthand.
While I applaud anyone who figures out a way to attend all 200, there are a few sessions that you should definitely see if you want ideas to help boost your team’s productivity.
One that comes to mind is the Building stronger teams with team-based functionality session. Think about when you work on a project at home. Now think about how you work on a project at work. Do you find that your work’s success depends on a team of people rather than one person? Most would say yes. Yet, historically, productivity tools have focused on helping individuals get more done — like how you manage your inbox or tackle your to-do list. Since we rely on teams to successfully complete tasks, we need tools to help that group be more productive as a whole. It’s a new concept, and I’m excited that this session will share some of the early work that we’re doing to move beyond individual productivity to, instead, use technology to help entire teams achieve more.
Businesses hear all the time about how machine learning can have a positive impact, and many are interested to see how they can achieve that same impact for their companies. Fortunately, Google has always been at the forefront of machine learning technologies like computer vision, predictive modeling, natural language processing and speech recognition.
To that end, I recommend checking out Machine learning powering the workforce: Explore in Google Docs to see how machine learning in G Suite can instantly help you tackle everyday tasks and complex business challenges with the click of a button. Then, follow that up with Introduction to Google Cloud Machine Learning to learn how you can build your very own custom business applications on Google Cloud Platform (GCP).
Whether it's using the Sheets API to give project managers using Asana a way to do deeper comparison of their projects, or using the Slides API to create a deck in Slides from a Trello board in just one click, the ways in which our customers and partners are automating their processes using G Suite APIs are impressive (and growing). The APIs we’re building across G Suite, as part of the larger Cloud platform, are being tailored to solve the most common business flows and the Automating internal processes using Apps Script and APIs for Docs editors session shows how some folks are already using Apps Script to make their internal processes hum.
These are the sessions that excite me, but you can find the sessions that excite you in the full Next '17 agenda. And if you’re wondering, you can still register. Grab your spot and I’ll see you there!
from The Official Google Blog http://my.onmedic.com/2lavBhK
Feb 7, 2017
The dark art of search engine optimization could be next in line for software-powered automation — potentially putting hundreds of thousands of ‘SEO consultants’ out of a job.
At least that’s the scenario sketched by RankScience, a YC-backed startup just graduating from accelerator’s winter 2017 program, whose software-as-a-service automates the process of running thousands of A/B tests in order to identify which changes will improve the Google ranking of customer webpages in organic search results.
“Ultimately that’s who we do see ourselves replacing,” says founder Ryan Bedner of the humble human SEO consultant who spends their days eyeballing scores of dashboards to try to identify beneficial tweaks. But whose days of gainful employment could be numbered if manual analytics can be overtaken by automation.
“It is an industry that I think we can disrupt,” he continues. “Humans are tweaking and measuring and changing, and software is really where things are going — and we think this is the start of that.”
So the pitch is: goodbye SEO dashboards and specialized in-house staff; and hello subscription software for automated testing and continuously optimized web pages. RankScience claims an average boost to organic search traffic of 37 per cent within three months, arguing such gains are a substantial step up from the competition — albeit it only has “a couple of dozen” customers at this point.
While Bedner says the SEO automation testing approach works well for “all sorts of different sites”, he flags particular benefits for ecommerce sites, marketplaces, directories, Q&A pages — “anything where companies have lots of pages”.
“Our software identifies areas of opportunity and things that companies should be testing based on where they’re ranking now, what they’re competitors are doing, and what opportunities we see. And also this data that we have from across our network — things that we see that are working elsewhere on similar sites,” he says.
The platform soft-launched in May last year, and the team has not yet done any active marketing. Most of the customers thus far are US-based but it does have some as far afield as Taiwan.
RankScience’s method requires customers to route their web traffic through its CDN in order that it can run thousands of concurrent split tests on their behalf, although it describes the set up for this as very easy; “two minutes” and a simple “one-line code change”, is the claim.
It also claims it’s doing things differently vs most of the startup competition in the SEO space because it’s not just doing analytics; it’s also automating making the changes too — taking a further piece of search optimization hassle out of its customers’ hands (assuming, of course, they’re comfortable relinquishing a little control over how their online content is structured, though it sends weekly reports about changes to keep customers in the loop).
The “most similar” competitor Bedner can be coaxed to name is BloomReach, which does ecommerce SEO for Fortune 500 companies, but he adds: “As far as I know, our continuous, automated split testing software is something they’re not doing. They don’t have a CDN, they do hosting for companies, so I think what we’re doing is differentiated from what they’re doing.”
Bedner says the idea for automating search engine optimization came from his “previous life” as an SEO consultant. “I realized my main advantage, related to other consultants, was that I was a programmer, and so companies could add me to GitHub, and instead of just sending them PDFs with recommended changes I could actually execute the changes myself… Our CDN is an attempt to productize that,” he tells TechCrunch.
“Almost all of the other SEO software products are analytics tools. They give you insights into how you’re doing with rankings, or they maybe make recommendations around things that you should change,” he adds.
“Our product is the only piece of SEO software that actually does work for you. So instead of creating tasks for engineers or product managers, our software actually handles the work for you, and executes for you. Because we’re a CDN — we can actually make changes to your pages. And other products can’t.”
RankScience not only carries out A/B tests for its customers, it also identifies which SEO experiments to run — although at this point it’s not yet fully automated that part of its process. So it remains to be seen how they can scale that, and what impact fewer “human inputs”, as Bedner puts it, will have on the results it can deliver for customers.
“For customers it’s 100 per cent completely automated. On our end it’s mostly software with some human inputs,” he says of the product at this point. “When a company comes on board we do interview them to learn about things they care about etc, so there are some human inputs as well. But the rest is software.
“We identify what to change and the larger our network gets the more powerful the data we have becomes because we know what’s working on other sites across our network — and so we have a good idea of what SEO experiments companies should be running.”
Bedner emphasizes its SEO methods are bona fide ‘White Hat’. So there are no dubious techniques being deployed in a bid to boost search visibility; it’s entirely compliant with Google’s best practices, he says, describing himself as “totally confident” the approach won’t get a customer’s website penalized by Google for trying to game its algorithm.
As a consequence of looping traffic through RankScience’s CDN there is a very small amount of latency involved in the process — “typically around 16m/s of latency or less” — minimal on account of how many websites are now hosted on Amazon Web Services’ cloud service platform, according to Bedner.
“The reason it’s so fast is that most companies now are on AWS and we spin up our CDNs in the same region as companies’ origin web server’s on AWS. That’s one of the reasons that this could’ve never been built before. Now everyone’s basically in the same data center — AWS enables this speed.”
And if customers are using other cloud service platforms — say Google Cloud or Microsoft Azure — he says RankScience can just “spin up our CDNs in the same region”, so the latency will never exceed 16m/s regardless of the platform a customer is using.
The team of five has raised a small pre-seed round from friends and the founders own money. Their next step on graduating from YC will be raising a seed.
“We’re totally focused on scaling right now. We’ve spent this last year focusing on product, and validating a lot of our hypothesis. The past six months we’ve just been trying to add more customers, learn more about how this works for different sites and how people are using this. So we’re just totally focusing on scaling at this point, and improving the product,” he adds.Featured Image: Global Panorama/Flickr UNDER A CC BY 2.0 LICENSE
from TechCrunch http://my.onmedic.com/2jXMQ4y