Online Positioning: Picking a Keyword Mix


It seems like every 5 minutes a blogger writes about how to do SEO.

But there are only scarce fragments and side notes on how to actually research and craft a comprehensive strategy & plan for how SEO will be done most effectively for a particular website.  This is staggering.  The competition online is simply too fierce, too large and too varied to risk compromising on strategy.  In this post, I’ll provide a few of my own SEO mental models and some concrete guidance for those who want to get their SEO efforts right the first timeIt’s not rocket science, but it does take some discipline, attention to detail and analysis. 

So here we go.

First, the 80-20 rule.  Ideally about 20% of your time allocated to SEO is spent on prep/strategy, while about 80% is spent actually doing the SEO – creating & optimizing content, improving site structure and building external links.  But from my experience, roughly 80% of success in SEO is the result of effective strategy and planning.  For one particular site that required about a month of SEO research, organic traffic had increased 100% within just 3 months of the initial roll-out of the 1st phase of content/structure changes. And don’t get me wrong – the site was already ranked fairly well across a variety of terms.  It just hadn’t picked the right terms – it was not yet positioned hierarchically or around the core value prop. or target audience.

Research, Analysis & Keyword Selection

SEO Process Cycle

Differentiation
Just like any marketing plan, you need a competitive position online that is based on your specific value proposition and core audience.  You can’t say “me too” online for very long.  There’s far too much of that going on.  The key to SEO positioning is centering your keyword mix on the differentiating factors of your value proposition.  You need differentiation to have a position that attracts quality traffic, but you also need to test these keywords.  This is especially true for smaller sites and smaller businesses.  Understand who your competitors are online and study their ranking and website content to ascertain where they’re strongly positioned and where they’re vulnerable.

High Quality Keyword Mix
Jump into Google Analytics and look at organic traffic and conversion rates by keyword.  Pick the top performing candidates. If you don’t have historical conversion data associated with particular keywords, a modest AdWords budget is an ideal channel for discovering how to refine keyword mix and gain some quantitative support to the keyword mix in consideration.

Highest Possible Search Volume for Keyword Mix
Using keyword traffic research tools, you want to pick the most popular keyword variants within your mix of candidates to be the focal point.  Use AdWords keyword research tool and Google Trends. The operative word here is highest possible – don’t sacrifice quality for quantity unless you plan to be a gigantic traffic eating monster like Wikipedia or eHow.

Long-Tail
As one hypothetical example, it is possible to take a word like “craft beer” and turn it into longer, more specific keywords, like “seasonal craft beers” and “IPA craft beers”.  In this way, the core of your focus is around a broader, high-volume keyword mix; but you extend and diversify your keyword array around those core terms.  This makes it easier to rank high for keyword variants that competitors may not have thought of.  The nice thing about longer keyword phrases is that they tend to produce higher conversion rates.

Keyword Strategy: Visualizing Structure, Defining Variants,  Applying Keywords to Site

The Plan

After the keyword mix is visualized and clearly defined, it is crucial to have a concrete plan and timeline for executing an SEO strategy.  The first phase is site/content optimization.  The second phase is content creation (blogging would be an ideal example of this for small businesses).  The third phase is external link building.  All of these can be done concurrently, but it is important to at least emphasize each of these phases in the stated order.

The final component of the SEO plan is a roadmap for measuring and improving upon the SEO strategy over time.  If you can’t quantify the impact of an SEO strategy, it’s a waste of time and money.

What to measure:

  • Rank is still a useful leading indicator; however, don’t forget that there are thousands of keyword variants used to find your site.  So even if you don’t rank that well for your core keyword, you could be taking #1 spots for lower traffic, higher quality terms.
  • Traffic – Relative and absolute growth month to month of organic traffic
  • Conversion Rate of organic traffic high-level and broken down by keyword.  At the very least, track registrations (or the rough equivalent of that for your site) and transactions (or, if revenue is not your business objective online, whatever most desired action defines the whole of your site)
  • Retention & Customer Value $ (Advanced) – You need a database/eCRM for this.  But it is extremely helpful for quantifying the outcome of SEO efforts. If you don’t have a database or eCRM solution, there a good make-shift solution: take the number of new user registrations by month (even if that is a simple .xls file of email addresses w/ dates) and subtract your Google registration conversions.  Then, apply your AOV or average annualized customer value $ to the share of new registrants that acquired through organic search. By taking these two simple steps, you are now at least able to quantify the share of new users acquired organically through unpaid search and the approximate share of annual revenue those new users produced.

A few pointers on how to measure and analyze effectively:

  • Annotate EVERYTHING – every major change to site structure/content, every ranking milestone, every major Google algorithm update, every major external link building effort.  I guarantee you’ll need to look back in order to do effective analysis and be able to articulate the story of your SEO growth.  There are far too many variables to keep track of in your head.
  • Use Rank Checker to document rank improvement over time.  Later on, you’ll want to overlay this data with traffic & conversion data from organic traffic.  You can automate this Firefox add-on to run routinely and check a long list of keywords.  Make sure you’re logged out of Google, etc. when you run it or your rankings will be skewed to your personal search history.
  • Everyone can afford to set up a free Google Analytics account and pay a web admin or developer to add Google Goal TrackingGoogle Webmaster tools is an important insight tool – I particularly appreciate the previous month comparison chart for keywords.  It shows you how much you’re up or down for any given keyword (traffic, click-throughs, etc.).  Webmaster tools also provides html suggestions, which is helpful for those just getting started in SEO.
  • Use AdWords keyword research tool.  It’s still one of the best tools.  While your at it, setup a modest AdWords campaign ($500) and get some initial conversion data on different keywords you’re considering.  Although, make sure your statistical sample (number of click-throughs) is large enough to actual warrant any serious conclusions.
  • Review & Analyze weekly, monthly, quarterly & annually – experiment with SWOT analysis approaches and document everything you find.  When you annotate your changes, try to include a concise summary of the analytical justification for the change as well.

That’s about it.  If you follow these steps at the outset, you’ll have a clear sense of where you’re going, you’ll know where to get started and how to prioritize the phases of execution, you’ll know what to do when things aren’t going well, you’ll be able to explain to others why things are going well, and you’ll know how to continually build on the growth and momentum you’ve started.

I highly recommend SEOmoz, SEO Book and “The Art of SEO“.  Great resources for the beginner or aspiring expert.  These thought leaders provide far more insight and expertise than I do on the subject and I couldn’t have had success without them.

Advertisements

Online Performance Measurement in a Nutshell


1. What is website engagement?
Website engagement is the combination of the quality, quantity, frequency and depth/complexity of user actions/interactions, outcomes & experiences on a website that matter to that website’s owner.

2. How should it be measured?
Website engagement is measured by qualitative and quantitative KPIs that track how much (volume/absolute metrics), how efficiently (ratios/rates of engagement/ROI) and how exceptionally (customer satisfaction index, etc.) a website is delivering on its promise to its target audience(s).

3. What specific approach and measurements should be used?
From a high level, it involves the whole customer lifecycle, measuring interactions/conversions at different phases in the customer-company relationship, from the point of registration/first use, to first purchase, repeat purchase and other engagements related to long-term retention.

From a ground level, I would look at micro-conversions that logically and actually prove to be leading indicators of growth in the area of key business outcomes. I measure engagement on the current website I manage by tracking a closely linked series of metrics.

First, I look at registration conversion rates and bounce rates to gauge the quality of the traffic I’m driving to my site. Then I look at the absolute volume of new traffic I’m driving to my site, in order to track the reach/exposure of my organic/unpaid and paid advertising efforts, as well as other marketing channels (email, SMS, RSS, etc.).

After that, I turn to site usage data, like entry & exit points, keyword & referring site data, avg. # of page views, common internal searches, etc. Then I look at conversion rates, CPA (cost per acquisition/conversion) & ROI, and the average window of time it takes for a customer to make their first transaction or key conversion.

Then I look at overall conversion rate for that first transaction. After that, I evaluate retention related metrics, like % of customers with more than X number of transactions and % of customers with and AOV greater than $XXXX. (This assumes a few things about the given business model in which I am operating, but it is for the most part broadly applicable).

Lastly, I would gather survey data and call center data to measure customer satisfaction and gather insights into customer behavior on the website.  If you can manage it, talk to customers directly or reach out to them on customer complaint forums or by email.  Any channel you can use to extract valuable qualitative data is priceless.

Phew! That was a mouthful.  Hopefully this helps frame website performance in a holistic light, if nothing else.  These questions were raised in a course of mine and I thought they were simple, but pointed.

Below is a model which I developed to frame online performance & engagement measurement for an auction e-commerce website I oversee.  It visualizes website engagement KPIs categorically, from the outside-in, moving from site usage data to business outcomes to actionable insights for the core of the business:

website engagement measurement

This graph is loosely based on Avinash Kaushik’s model in Web Analytics 2.0.

Is the Funnel Really Dead?


I am a firm believer in lifecycle marketing.  Thinking of the entire lifespan of your customer challenges you to revisit your brand strategy and build more robust marketing tactics that extend beyond merely transactional touch-points and improve overall user experience.

But recently Performable, a web marketing tool provider, has made an interesting claim:  “The Funnel is Wrong”.  On his Forrester blog, Steven Noble goes as far as to say that we should bury the conversion funnel altogether.

The Performable whitepaper says there are a lot of things wrong with funnel marketing, but the bottom line is: funnel marketing doesn’t reflect customer behavior, tends to separate marketers from customers, doesn’t provide enough context for the user experience and doesn’t account for a particular phase in the customer lifecycle.  The outcome of the funnel is a simplistic approach to optimizing web marketing efforts that is not nearly as holistic or effective in the short-term or long-term as lifecycle marketing.  Read whitepaper: Why Lifecycle Marketing is the Future

Here’s how Steven Noble illustrates the comparison between the funnel and lifecycle frameworks:

Funnel Vs Lifecycle

You get the basic idea – one depicts a linear path that assumes each transaction results in increased loyalty.  The second involves a variety of experiences with a brand that gradually and cumulatively drive not only more future transactions, but a personal loyalty to the brand.

Apples to Oranges

First of all, I am not sure how funnel marketing is opposed or even comparable to lifecycle marketing.  From my experience, funnel marketing is a tactic that can easily be crafted to fit within a lifecycle marketing strategy.  There might be plenty of ways to terribly execute a funnel marketing plan and to execute such a plan without any consideration of strategy, but that doesn’t at all indict funnel marketing per se.

Comparing lifecycle marketing and funnel marketing is like comparing Hitler’s strategy for dominating all of Europe with General MacArthur’s island-hopping tactics at Okinawa.  It’s just not comparing apples to apples.

Evaluated as a strategy, funnel marketing would look pretty ridiculous.  But that is because it isn’t a strategy at all! It’s really easy to shoot down the weaknesses of a funnel-based approach if you’re thinking of it as a strategy.

An Officer, Not a General

Yes, conversion funnels (on their own) are wooden, overly linear, too focused on transactions and don’t account for the variety and depth of user experiences. But has anyone said funnels were supposed to more than this? Perhaps someone has said it, but I don’t know of a single reference to funnel marketing as a comprehensive online marketing strategy.

In any case, I do think Performable could make a point. It seems like the real enemy isn’t the activity of funnel marketing, it’s narrow-mindedness that focuses merely on transactional conversions, without the broader context of a lifecycle strategy to inform other critical touchpoints and non-transactional conversions/actions that are needed to cultivate a long-term relationship with customers.

KPIs and conversion funnels are narrow-minded on purpose they have to be in order to make sense of data and adjust quickly.  But web analysts, good ones anyhow, know that the KPIs aren’t a substitute for an over-arching strategy.

You don’t let an officer direct an entire military force, but you still need him.  Understand his role, and he could be the difference between success and failure.

There is Still a Heartbeat

There are good reasons not to bury the conversion funnel. In the web intelligence/analytics community there is a strong emphasis on tracking KPIs at the end of funnels, and I think this is crucial for a few reasons:

  1. They’re quantitative
  2. They’re based on business outcomes
  3. They provide a focused and simplified view of web performance that enables quick analysis and action

#3 is important because of a few huge problems with how web analysis has been done historically: Too much data, too many reports, tons of data that may be interesting but only 1% or less that is relevant or actionable.

Funnel to Freedom

So even if I sometimes get trapped in my conversion funnels, I force myself to periodically take a step back, look at the strategy, and make sense of the data within that high-level context. Ideally, that strategy includes the customer lifecycle. This is not only good common sense, but it is also psychologically refreshing for marketers like myself who thrive off of a sense of the big picture and how the end-user is actually benefiting.

Looking at data through the lens of the overall strategy means, of course, that my conversion funnel data should include conversions that are not just transactional, but also track other key interactions in the customer lifecycle, like survey completions, social media mentions, product recommendations, “share with a friend” actions, etc. In other words, my conversion funnel data can and should provide me with insights about the overall health of my customer lifecyle.  If I haven’t defined my KPIs around a unified strategy, how will they be of any use to me? In this respect, I disagree with the criticism that funnels can’t tell us about actual customer behavior or experience.

A Modest Proposal

I think Performable’s message needs qualification: funnel marketing is only wrong when employed without a strategy, which lifecycle marketing can provide.

As a web analyst, I know how incredibly easy it can be to get buried in the numbers and fixate on conversion funnels.  But I can’t throw the baby out with the bath water, or bury the baby with the bathwater…. hmmm, my imagery is becoming problematic.

There is a place for funnel marketing within customer lifecyle marketing. If you see yourself letting the funnel become the be-all end all, pull yourself out and revisit your customer lifecycle strategy.  Whatever you do, don’t pull the plug.

But that’s just my take at the moment.

Do you think there’s hope for funnel marketing or AIDA related models?

If you have a moment, make sure to check out Performable’s software.  It is quite innovative and ambitious in its attempt to gauge customer lifecycle performance. I haven’t seen a demo yet, but the concept is certainly intriguing.