Automated SEO Strategies
For many companies, SEO can be the “Front Door” — a blog post, or content page is generally the first interaction that a user has with a company’s web site, not the main page of the site. As such, the content of the SEO pages are important in understanding much of how successful a site will be in terms of conversion and retention.
The aim of this article is to outline several different strategies that startups and larger companies can use to grow their SEO in an automated or semi-automated fashion.
We begin with an overview of key points to consider in SEO strategy:
· Content Quality and “Thinness”
· Consumer Intent & Funnel Stage
· Competion for the query type
Content Quality and “Thinness”
Google penalizes content that is “Thin”. “Google will take action against domains that try to rank more highly by just showing scraped or other cookie-cutter pages that don’t add substantial value to users.” In particular, it calls out content that is “scraped”, i.e. taken from another source, where the new domain doesn’t add “original content or value.” So first, make sure that this content isn’t readily accessible to Google elsewhere — i.e. it’s in PDFs, images or is inside a site that is not crawled by Google. The fastest way to check this is to copy a paragraph of content and paste it into google and search for it. Google will show you all the sites where the content is being posted. Second, make sure that you are adding value. The most common way I’ve seen to add value is to organize the data, both by putting it into HTML, but also by codifying it in a standard way. So, for example, Yelp codifies all restaurant data into a series of attributes — type of cuisine, hours, good for groups, free wifi — and makes it easy to then search on those attributes from each programmatically generated page. The important point is to make sure that the attributes you pull from the data are important to the end consumer.
So, as a company, one great question to ask yourself is, what unique data do I have access to, or what is a great data source that no one is building logic on top of? More on this in the section below.
Scraping for Competitive Advantage
Some well known companys were built on scraping data from other sites and adding better search, ecommerce or analytics. One famous example is AirBnB, which scraped vacation home listings from Craigslist, made them easily searchable, and added in a reservation system. Many companys have begun in a similar fashion, finding a category on Craigslist and then building a much better user experience. Why is this a popular approach? With marketplaces, the first problem is usually getting supply. Scraping a directory provides a way to bootstrap supply initially, and then attract demand, which drives real suppliers to the platform. For marketplaces, the best way to apply this strategy is to start with a single city, so that it’s easy to reach critical mass. Besides marketplaces, analytics companys have also been built off scraping other sites.
An example of an analytics play is the company AirDNA, which scrapes listings from AirBnB to provide analysis tohosts about how different property types do in each city — how much additional rent does having an extra room create, what is the average monthly occupancy rate of a house vs a condo, etc. Another analysis play is Mattermark, which analyzes the twitter and linkedin profiles of companys to determine which ones are gaining traction.
So if a company is heavily reliant on scraping, it’s good to ask if the strategy sustainable and scalable. It’s best if it’s a one-time booster rocket, or the site being scraped or accessed via API is trying to court 3rd party developers. One common peril that leads to company death is the data source company removing an API or changing its data sharing policy. If the company will have an ongoing need for scraped data, it’s best if that data comes from several different sources, so that if one goes away, it won’t be too much of a setback. Mattermark, for example, uses Linkedin, Twitter, and Facebook in combination.
Consumer Intent & Funnel Stage
A more suble consideration is where the user is in the marketing funnel when he lands on the site. Content that is oriented around purchase-making decisions, such as lists of the 10 best products, X vs Y product etc. are said to be “Down Funnel”, the point in the marketing decision funnel when people are closest to making a purchase decision. This is based on the idea of a Marketing Funnel that starts at attention and moves towards purchase:
In general, content that gets into the specifics of products or service providers is the most down funnel. In particular, individual product pages, or product directories that list prices, are the most downfunnel type of content. “Upfunnel” content is typically more general opinion or Q&A. For examples of this, think of internet community forums, Facebook discussion or Quora or Yahoo questions. These types of content are typically used by a consumer doing research. While these pages can be great at showing up in search, they’re not very good for converting into a purchase, unless the page is designed with a strong call to action in mind.
Competion for the query type
From an SEO perspective, one of the first and most important decisions a company should make is what type of query type to go after (hat tip to SEO expert AJ Kohn of BlindFiveYearOld.com). A query type is a group of common search terms, usually related around a product or service. For local businesses, the query type is usually [CITY] [SERVICE]. For national companies, the query type is typically the name of the product and variants, e.g. “Gaming Laptop”, “Fast Gaming Laptop”, etc. For SaaS companies that serve a niche, the query is often the general name for the type of service or need it fulfills, such as “Marketing Automation” or “Product Management Software”.
A good proxy for the competitiveness of the market space is how competitive the query type is. A query that results in many ads is highly competitive. Also highly competitive is one where the first 10 results are all nationally recognized brands. Beyond that, there are subtler signals to show how competitive a query type is. Moz.com is one of the leading SEO training companies, and it offers an “SEO Toolbar” (Chrome plugin) that shows the “Domain Authority”, or overall search competitiveness, of the companies that show up in the Google results pages.
In the box below, the search query is “Mortgage Rates”. The Moz extension shows the domain authority as “DA” below each result, with a score from 0–100.
As we can see by looking at the picture, you may or may not have heard of brands like Bankrate or Nerdwallet, but they are highly competitive with better known brands like Wells Fargo, LendingTree and Zillow.
So, if a query on Google shows well known brands, or domains with high domain authority, and a company doesn’t yet rank, should the company go after that query? This is where it gets subtle. A few factors come into play:
In general, the more specific the query is, the easier it will be to compete for. “2015 Toyota Corolla for sale in San Francisco” is much easier to rank for than “Toyota Corolla”.
A second factor to consider is the “Page Authority”, as measured by Moz — How useful are the actual pages to the consumer. Do they have good original content, or is it just a few lines of text and a picture. Page Authority is measured by how many sites link to that specific page. In the picture above, page authority shows up on the left-hand side as “PA”. We can see that most of these pages have strong page authority, except for the one from Bankrate.com. The Bankrate page is showing up because the company, rather than the page, is well known.
A company that has similar content to existing content from an incumbent is generally not competitive, because the incumbent has so much more domain authority. What then to do? Your company should either have content that has much more depth, or should find a niche to dominate, or both.
Types of SEO
SEO breaks down into a few categories: programmatic, content, user-generated content (UGC) and tool-based.
Programmatic SEO is based on focusing on a search query type, and creating a strategy where software creates hundreds or thousands of pages to capture different variations of that search query. One common example is
[CITY] [SERVICE], e.g. San Francisco Restaurants. Yelp is an example of a company that executed extremely well on a programmatic strategy, not only having a programmatic page for every city, but every neighborhood in major cities, e.g. Soma Restaurants, Potrero Hill Restaurants, etc. Programmatic SEO traffic can be a great source of traffic for a company, and many companys have used this as their main strategy.
Tools like SpyFu and SEMRush will give you an indication of the query type that the company is focused on, as well as how competitive they are for their queries.
Another variant on the Programmatic strategy is to put public content online in a searchable format. For example, BuildZoom is a website that helps people find commercial contractors. Each contractor has a public record maintained by the city. These records are publically available but not designed for SEO — they may be PDFs, scanned documents etc. Buildzoom takes each contractor public record and creates an HTML document from it. They then leverage the built-in search volume for each contractor’s name to drive traffic to a page about that contractor showing their public record. The page then links back into BuildZoom’s main page.
Making public content searchable can be a great strategy, but it should build off of
existing search traffic.
“Content”-based SEO usually means blogging. A few general points on making content SEO effective. First, go for quality over quantity — if you post inane “marketing” content once a week, you likely won’t get the links you’re looking for.
Second, content blogging, and most forms of SEO, are a long game that won’t pay off for at least 6 months, and more likely 1 year or longer. If a company is pursuing this strategy, it will definitely take time to see a payoff. Content headlines should be relevant to the general business questions that the company’s customers have, rather than an ongoing pitch for the product. So Optimizely, a company about user testing, will post on A/B best practices, statistical tests for online testing, etc. Another popular method for finding ideas to post about is to look at popular search queries for your category, and use those as your headlines.
As a general rule, long-form is more valuable than short-form. A good general strategy for a company to pursue is “The Definitive Guide to X”, where X is a business skill that the company aims to support (e.g. user testing for Optimizely). Neil Patel / KissMetrics and Moz are great examples to look at of people who’ve succeeded with long-form guides. Note that their long-form guides are not behind a signup wall.
Instead, they just do a soft-sell for people to join their email lists to receive ongoing content.
User-Generated Content or UGC has been the mainstay of several successful companys, notably TripAdvisor, Quora and StackOverflow. UGC is even more of a long game than content, since the company has less control over production, taking at least a year to produce a noticable effect. One way around this is to hire writers to seed the User Generated Content, by posting questions and answers (under different accounts), and interacting with other questions that are posted. A second way to encourage more content is to build in social sharing, so that when users ask questions, they’re encouraged to post a link to them on Facebook or Twitter.
When a company is heavily dependent on UGC as a traffic strategy, it’s important to understand the motivation of the contributors. Two broad motivations are competition and community. A competitive motivation exists when the writer wants to show off their expertise. In a competitive case, users can become more motivated by Gamification. At a very basic level, gamification means using techniques like points (measuring number of contributions), badges (rewards for specific actions), and levels (rewards for total contributions, tied to points). These techniques help the writer show other people on the forum that he’s a recognized expert in some area. In a community motivation, by contrast, writers derive meaning from helping others, and recognizing others’ feelings. The best known example is in social media, the Facebook “like”, which lets writers communicate approval to each other. In this type of user-generated content, a company is best served by encouraging users to post frequently, even if they have nothing particularly new or important. Instead, it is the ongoing relationship that is important between users, not the particulars of the content itself. So a community / social-media type of company is better served in asking the user simple questions that prompt frequent engagement — e.g. How are you feeling today? What are you excited about right now? , or soliciting users opinions on a shared topic of interest, e.g. what do you think about the new iphone, etc.
Tool Based SEO
Tool-based SEO comes from a well designed or novel free tool that either gets publicity on its own, and / or gets SEO for its tool name.
An example is Paint My Photo [https://www.instapainting.com/ai-painter], that takes a user’s uploaded photo and converts it into the style of famous painters.
As another example, VisibleThread offers a free content clarity grader [http://www.claritygrader.com/] as an incentive for learning more about their proposal management software.
Tool-based SEO can be great in that it can be difficult to compete with, and long lasting. The downside is it tends not to scale up, and it’s hard to create a good tool in the first place.
A company that has only tool-based SEO will usually need to supplement it with another source. An exception is when the tool programmatically creates results pages that have high SEO value, such as a salary calculator that creates results pages for common professions and locations, eg. San Francisco Product Manager. In this case, the tool SEO may create enough traffic to serve as a companys main traffic source.
In this article, we’ve reviewed several different strategies you can use to boost your SEO at scale. If you’re interested in learning more, contact me at firstname.lastname@example.org.
For a consumer company, traffic numbers are a good proxy for interest and business. For a B2B, this is also true, but B2B numbers will usually be much smaller than B2C, so you should scale your expectations appropriately.
For consumer websites, use one or more of the following sources as a traffic research tool:
These sites generally get their data from a browser bar, or buying data from ISPs.
So the absolute numbers are usually off, sometimes by a large factor, but they are good relative indicators; Benchmarking companies against their more established competitors gives a good general indicator of how a company is doing.
Beyond sheer traffic, what else should you look for? The growth rate.