How Your Keyword Strategy is Failing You (And How to Fix It)

One day very soon, we’re all going to stop worrying about “keywords”.

And then we can all rejoice.

via GIPHY

But until that happens, whenever “SEO” comes up, the next phrase muttered usually involves “keywords” and “rankings”.

This unhealthy obsession with “keyword + rankings” (that was a search operator joke) has long been misleading.

Historically, keywords gave us a fairly reliable way to measure progress in the otherwise abstract and confusing world of SEO. (Not to mention, the very real danger of cheap SEO providers.)

The problem is that today, keyword rankings are basically useless. Which means the way we’ve traditionally optimized and measured for them is basically useless as well.

Here’s why.

Why Your 2005 Keyword Strategy Doesn’t Work in 2015

Rightfully (or wrongfully), keyword rankings have been SEO KPI #1 for over a decade.

And back in the day, this made sense.

Search engines were more-or-less one dimensional, which made SEO a very straightforward process. Everyone, no matter who you were or where in the world you were searching from, would largely see the same exact search engine result pages (SERPs) when looking for a specific keyword.

In this environment, keyword rankings (as a metric) were very simple, reliable, and (I can’t believe I’m going to say this when referencing Google) transparent.

On top of that, analytics programs freely passed keyword referral data back to webmasters. Meaning you could see exactly which terms people used to arrive at your site.

By matching keyword positions or rankings with the referral data you were seeing in analytics, you could easily see the $money keywords – or which ones were driving success (in terms of traffic and conversions).

Unfortunately, none of this is true anymore.

So good thing you scanned over the last ~165 words anyway. :)

Something About Personalization

Today, everyone’s search engine result pages (SERPs) are being personalized based on your:

1. Past browsing history
2. Physical location
3. Social media connections

Just to name a few. :/

That means the keyword rankings you’re seeing, instead of being static and universal like the good-ol-days, are completely personalized to you as an individual.

For example, look up a traditionally head (or super popular) keyphrase like, “Pizza” and you now get this:

Modern SERP’s pull from a variety of different sources (here you’ll see the huge prevalence and opportunity of local search emphasized), with traditional “organic” results pushed off a bit.

(This also means the role of “SEO” has evolved to include influential satellites like AdWords and Yelp. But that’s a topic for another day.)

The concept of keywords having one specific rank, and then benchmarking efforts against it, is today at best worthless, and at worst misleading.

But wait, there’s more!

Dude, Where’s My Keywords?

The second part of the keyword ranking equation was using referral data from your analytics to see how and where people are coming from.

With this info, then you could at least get an idea of (a) how people are looking for you and (b) how to use that information to do a better job of optimizing your site.

So even if keyword rankings are losing value, this referral data was extremely helpful in giving you clues to influential topics and keyphrases.

Now, SEO encompasses much more than just Google Search. Hoooooweeevverrr… Google Search is a virtual monopoly, meaning they can pretty much do whatever they’d like. Starting with, taking away almost all keyword referral data that gets passed to webmasters and site owners.

A few years ago, they moved to make all searches secure (except for ad clicks). Now in your analytics program, where you used to see the specific keywords sending you traffic under “organic”, you now see a [not provided] placeholder that accounts for the majority (~70-90%).

That means you can no longer see what keywords are sending you traffic from organic search….

… due to “privacy reasons”…

… but you can, however, pay them for it via AdWords.

How ironic. And convenient.

Keyphrases are still very important. Trouble is, we now have to infer or assume what keyphrases are popular and how to best optimize with huge gaps in verifiable data (and you know what they say about when you assume).

One of the easiest ways, is to simply alter our strategy a bit and focus on what we can control (our website) instead of what we can’t (keyword rankings).

The Simple Change to Update Your Strategy

If (a) keyword rankings are unreliable, and (b) keyword referral data is nonexistent, then… something needs to give.

Going forward, it’s easier to shift focus away from keywords (directly), to the performance of your landing and content pages instead (so you can indirectly assess topic performance).

Then reverse-engineer success based on topic – i.e. a broad set of long-tail keyphrases – instead of only one specific keyphrase. It’s messy, but practically easier (unless you’re interested in getting your PhD in SEO and analytics).

For example, one simple way is to take a look at your most popular content in Google Analytics from organic search:

Then cross-reference this information with some (remaining) query data in Google’s Search Console (formerly Webmaster Tools):

And you can kinda get an idea of the long-tail keyphrases sending this page traffic (along with some position-related info – but let’s not over-emphasize this now, shall we?).

You can also use some paid tools, like Moz, to help track a certain number of keyphrases against specific landing pages:

In a way, this backwards process should actually benefit you by ensuring extra attention-to-detail when strategizing the information architecture of a site’s pages (and their respective keyphrases) in the first place.

The Holistic Future of Search Optimization

In today’s dynamic marketing landscape, SEO isn’t “SEO”.

Instead, SEO now takes a multi-faceted approach where you’re involving different disciplines (i.e. content, email, advertising, social), building a brand (i.e. investing in intangibles, not just conversions), and competing on multiple fronts (i.e. paid search positions, review & aggregation sites like Yelp, TripAdvisor, beefing up your local listings, and more) – all at the same time.

Needless to say, this requires a lot of time, man (or woman – can’t accuse me of discrimination!) power, and sufficient investment.

The days of competing solely on (and overprioritizing) SEO are numbered.

But that’s not to say it’s any less important. In fact, search is only becoming more important and more influential in the buying process of customers.

Finding what you’re specifically looking for will always be priority #1 online. And that means search will be omnipresent and omnipotent because it’s so valuable (and profitable).

The trick will be to remain holistic and nimble as trends and platforms evolve.

About the Author: Brad Smith is the author of a BS-free SEO guide that shows you how to fix common mistakes while avoiding algorithm chasing. He’s a founding partner at Codeless Interactive, a digital marketing firm digital agency specializing in creating personalized customer experiences.

Using Kissmetrics to Find Out If LiveChat Is Affecting Conversions

Live chat. If you don’t already have it installed on your website, you’ve undoubtedly spent hours contemplating and debating with your colleagues. But does it actually work at increasing conversions? If you don’t have an analytics tool that provides this data, you’re left in the dark. Kissmetrics is here to help.

Kissmetrics includes an integration with LiveChat. This integration can give a lot of different details; such as if/when a person sees a pre-chat survey, if it is filled out, when a chat is started, etc. Today, we’ll focus on using Kissmetrics to tell if customers that use LiveChat have a higher likelihood of converting. Of course, you will have to be using LiveChat to get any data.

Does Chat Increase Conversions?

Each business is different. I’m not going to make a blanket statement like “chat doesn’t increase conversions”. What works for some companies may not work for others. Since you have Kissmetrics, you’ll know what works for your company. Here’s how:

Using the Kissmetrics Metrics Dashboard, create a new metric. Select the option Number of People Who Did Event.

Enter the name of your new metric and select the event you’ll be tracking. If you’re in SaaS, it may be something like sign up or demo requested. If you’re in eCommerce, it would be purchased or created account. Remember, we’re tracking conversions here, and you’re using LiveChat to see if it moves some needle. The needle you’re trying to move is the event you’re tracking.

For filter property, select Performed Chat. This will automatically be in your account once you connect LiveChat to Kissmetrics. Save your metric and you’ll get the data.

And you’ll get your data. True means they did chat, None means they did not. What you see below is just hypothetical data. Your results will vary.

If you see that there is little effect, don’t give up on LiveChat yet. Here are some ideas of what to test in LiveChat:

• Change your greetings – LiveChat allows automated greetings to popup before a visitor even types a word into the chat box. Do some automated greetings get better responses than others?
• Compare chat agents – You may have 5 people that handle LiveChat. Does any one of them convert visitors better? With this integration, you can see the number of conversions each agent has delivered.
• View the chat transcript – This one is a little more high-level and requires some manual work. But if you’re really curious, you can go back and view each transcript for the ones that converted and the ones that didn’t. The data is all within Kissmetrics. Here’s what you’ll need to do:

Use People Search to Find People Who Fit a Specified Criteria

You’ll want to find the people that have done the event you’re using in the Metric. This can be signed up, purchased, etc. Next, find the people that have the property Performed Chat and it should contain true. You can make the date range whatever you’d like.

With this, you’ll find the people that signed up (or purchased, created account, started trial, etc) and have chatted. Click Search and get your list of people.

Once you see the people, you can go back into LiveChat and see the chat transcripts. How did the agent manage to get the person to convert?

Compare these against the people that didn’t convert. To do that, just change the Have done event to Have not done event. This will show the people that have chatted but have not converted. You can then compare the transcripts of those that converted and didn’t. Compare the language and conversation and how they were addressed during the chat. You may be able to pick up some good nuggets on what triggers conversion and what doesn’t.

Recap

Here’s the takeaways:

• Using chat on websites is a debated topic. Does it have any effect on conversions? Each website is different, and you’ll need to test it to see what effect it has for you.
• Kissmetrics includes an integration with LiveChat.
• Using Kissmetrics, you can see what effect LiveChat has on conversions. You’ll be able to compare conversions for people that used LiveChat and those that didn’t.
• Go beyond just seeing if chat increased conversions. Look at chat transcripts, test different greetings, or see if one person on chat is better at converting users. You can also view chat transcripts to see if there are any trends in what questions people commonly ask.
• If you’re already using Kissmetrics and LiveChat, you can view the instructions for integrating the two.

Questions? Feedback? Want to say hi? Put it in the comments.

About the Author: Zach Bulygo (Twitter) is the Blog Manager for Kissmetrics.

Make Personalization Work by Starting with the Right Data

Whenever I have the opportunity to chat with a thought leader in the marketing industry, I always like to ask: What trend do you think will be next year’s buzzword?

In 2014, many professionals said 2015 would be the year of personalization — a prediction that, as it turns out, may have been a little too hopeful.

There are a few reasons personalization hasn’t taken off as well as anticipated, but not one of these reasons is a reluctance to personalize. It’s not a matter of failing to grasp the merits of creating more relevant and better-targeted messaging, but a matter of mastering everyone’s other favorite buzzword and slated 2016 MVP: big data. Or, more specifically, obtaining, organizing and applying the exabytes of consumer data currently up for grabs all across the World Wide Web.

Of course, it makes sense that the year expected to be the “year of personalization” would be followed by a year hyper-focused on data. If personalization is a burrito, data is the rice, beans, guac … (BRB — going to Chipotle.)

What I’m trying to say is: You simply can’t have one without the other. But how do you get the right data to personalize in a way that engages your specific audience? Furthermore, where and how do you obtain and apply this data?

To get answers, I spoke with Phil Davis, Chief Business Officer and data and personalization guru for email data service provider TowerData. Here are his answers to the most common data questions.

How Do You Determine the Right Data?

The first step in achieving personalization is determining what data you need. While the sheer amount of data available to marketers is overwhelming, the good news is you only need to concern yourself with a very small subset of that information.

“The first thing you need to define is what you have access to,” says Phil. “Start with what’s at your fingertips: first-party data, and then third-party data.”

For most marketers, this breaks down into three types of available data.

• Behavioral data: The products or services customers purchase, and what products they view. For B2C companies, this can also mean looking at the contents of abandoned carts, and for B2B companies, this can mean considering which pieces of content a prospect viewed before becoming a customer.
• Brand interaction: What pages are prospects viewing on your website, and how are they engaging with your email messages? How did they reach your site in the first place? Do they follow you on social media?
• Demographic data: From age, gender and geographic location to marital status, children in the household, income and homeowner status, demographic data tells businesses a lot about the people behind the purchases.

“Get the data first, then figure out what data moves the needle,” says Phil. In other words, collect all the data immediately available, and then decide what’s most important or most relevant.

But don’t feel defeated if you can’t immediately obtain all the data you need for a highly sophisticated Amazon-esque personalization strategy right off the bat. As Phil points out, every business must start somewhere.

“If you can't get your hands on all of it, or if it’s too difficult to digest, just start with a few simple fields to get yourself on track for personalization.”

After all, if you jump from zero personalization to even a small amount of personalization, you’re going to notice an increase in conversions.

Where Do You Get Data?

“Oh yeah, big data is great,” we all say, collectively smiling and nodding. Here’s the dirty little secret marketers are hiding from one another: Most of us have no idea where to get all this amazing data we hear so much about. Sure, we may know who viewed which blog post and who downloaded which eBook. But what about all of those in-depth demographic data points mentioned above? After all, there’s only so many personal questions you can ask in a submission form.

The best option is to leverage the data you already have (first-party data) to get the data you need (third-party data).

Using tools like InstantData, businesses can upload their contact lists, choose from a giant list of data points and instantly enhance customer profiles. Marketers can add demographic, interest and purchase data to their customer database and voila! You have all the juice you need to power an uber-targeted personalization strategy.

How Do You Apply Your New Data?

You’ve taken stock of what’s already available in your database, and obtained even more data from a third-party source. Your arsenal is full, but how do you make sure you hit the target?

According to Phil, while the medium and methods of obtaining data may have changed, the application has not. “The machine is doing what a really great in-store salesperson would do,” he says. Only, the “machine” is more accurate.

“If it was 1970 and I was selling clothes and someone came into my shop, I would assume, ‘This person is in this neighborhood, so they must live close by,’” says Phil. “I’d notice he’s a man, he's about 45 and he's got a wedding band on. I'd probably ask him, ‘What are you looking for today?’ That's really no different than me wondering, ‘What did you click on today?’ and using data to provide a better experience.”

The Next Steps

While 2015 may not have been personalization’s banner year, the forecast for 2016 looks much more promising. As automation advances, and marketers become more comfortable diving into customer data, the ability to create highly targeted and captivating campaigns is becoming a real possibility for companies of all sizes.

Even the smallest amount of personalization based on a mere one or two data points can help marketers make waves. To learn more about how you can start analyzing your data for personalization, check out TowerData's InstantData tool.

The Top 5 Reasons You Should Start an Ecommerce Business

Even as a beginner in ecommerce, the tools and resources are all laid out for you. All you need is a strong work ethic and a desire to succeed.