“It’s too accurate” (more undocumented uses of NextStage’s Evolution Technology)

This post is about looking in one’s mirror and dealing with what is seen. This post’s origin is being told that the reason a company will not use NextStage’s tools is because the tools are “…too accurate.”

First — and I suppose it truly is a first — note that one of the owners of a company is sharing a reason a prospect won’t use that company’s product.

Second, I’ve actually glommed comments from a few folks into this post.

Third, because it’s too accurate???

I did thank the company for their interest, explained that we could always do business in the future, so on and so forth.

Then I hung up the phone and went back to wondering “…because it’s too accurate“?

Let me clarify this a bit. I’m honored by their decision, specifically the reasons behind it. This company’s principals were declining because they were, indeed, principled, and in a way NextStage can completely understand; before they resold our tools they would use the tools on their own material.

But there was some fear in their voice when they said, “Your technology is excellent. Nobody questions its accuracy anymore. You’ve published enough, others have published enough, it shows up in scientific material, Chris Berry even told everybody at his eMetrics Toronto presentation that they should go with NextStage if they want scientifically provable and actionable results, so nobody questions whether or not NextStage tools are accurate anymore.”

(thank goodness, that! And thank you, Chris, for that)

And then the kicker came, “We’re afraid to find out we’re full of BlueSky…” (they used another term) “…or something worse, like our designs really do suck and we always knew they did but could never admit it to ourselves. If we use your tools then we’ll have no choice but to face the facts.”1


We are sticklers for accuracy here at NextStage. It comes from the research background. It also appears in how we market (we’ve historically relied on word-of-mouth completely). We prefer to quote others who think NextStage is a hot-patootie than to say “NextStage is a hot-patootie” because we have a self-interest and therefore, from a research perspective, are violate on the subject (our opinions don’t matter). Someone else, someone with no interest in NextStage other than their own belief and experience?

Yippee, Great, Loving It and Go For It!2

As for tool accuracy? A market researcher who uses NextStage’s Sentiment Analysis Tool regularly to determine which companies are worth watching said, “Most people will use a tool that’s 80% accurate because there’s still 20% wiggle-room. Wiggle room means it can still be the tool’s fault if something goes wrong. But 98% or 99% accuracy? There’s no where to go with that and so far the tools haven’t been wrong, so now my feet are in the fire if I make a wrong decision. There’s not a lot of people willing to do that.”

Shades of accountability!

No One Can Look Into the Face of God and Live

The above line goes back well into antiquity and is (in various forms) found in oral and written traditions worldwide. The real idea is “No one can look into one’s self and live” and stated that way is at the heart of every shamanic culture and practice, every psychologic intervention and healing, everywhere in the world since such things began. The catholization of Europe moved the concept from looking into one’s own reflection to looking into the face of God, and the whole concept of self-exploration (quite accepted in the East) became gilded as “narcissism” in the West. 3

(Relatively) Recently the concept took the form of scrying mirrors. Scrying deals with divination and scrying mirrors with self-divination. Historically such devices were terrifying (The Portrait of Dorian Gray, wherein the protagonist stays as he wishes to be until he gazes upon his reflection and becomes what he truly is, is a literary example of scrying mirrors) because they deal with the (very real) belief that there is nothing more frightening and debilitating to one’s psyche than to be honest with one’s self. That “Knowledge of Good and Evil” thing? Study the original languages, study what different cultures meant by “The Book of Knowledge” and it always comes down to “knowledge of one’s self”.4

We at NextStage regularly go on retreats to perform such scrying. It’s not always easy…heck, it’s rarely easy. And I’m not talking as in “Web Analytics is Hard”, I’m talking as in psychologically, emotionally, spiritually, intellectually dealing with the real who of who you are.5

Are there aspects about yourself you don’t like and wish you could change? Welcome to the first stop in your tour de vous. Find out why those aspects “can’t be changed”. What can you do to change them in such a way that the change is 1) obvious to you then 2) obvious to others?

And you do this until you’re raw. I mean raw like you can’t have people around you because you’re so ashamed of the who that you’ve been and at the same time you desperately want those around you to tell you they love you all the more for what you’ve just put yourself through, that they’ll help you because you made the start, …

Now realize you’re doing this on your own, there is nobody around you to tell you they love you and will be there for you and will help you…

So that means you’ll have to tell yourself you love yourself, that you’ll be there for you, that you’ll help you, and remember that you’ve just emotionally, psychologically, spiritually, intellectually crippled yourself by going into that part of yourself you didn’t like in the first place and discovered it was there for an extremely, righteously, honest-to-god good reason that might not be such a good reason anymore, so maybe you should lop that part off, like cutting off a hand because it’s infected and the infection is spreading.

So that part of you dies. You’re raw because you just killed a part of yourself.

But you did it and in doing it, in recognizing there were things that only you could change, you gave birth to a new self, a new you and you’ve become your own phoenix rising from your own flames (people who’ve read Reading Virtual Minds Volume I, this is the Core work we were talking about).

And a NextStageologist requirement is putting yourself through such scryings regularly because it’s the only way we’ve found we can do what we do and be honest about it. Such scryings are what allow us to know how people are thinking without judging their thoughts.

Because the moment we recognize that we’re judging their thoughts we have to ask, “What is there about this information that’s causing me to become judgmental?” and sha-bang sha-boomie we pull ourselves out of work and put ourselves through another bout of scrying, another round of peeling layers off our onions and healing.6

So “Accuracy”…

Being told that a company doesn’t want to use our technology because they’re not ready to look into the mirror is completely understandable to us.

Very much so.

In fact, before we publicly released the NextStage Sentiment Analysis tool, we had a long conversation about accuracy, as in “Joseph, you’ve put a lot of information out there. At some point, someone’s going to analyze your stuff. You ready for that?”

The suggestion was even made that we teach ET to recognize my writing regardless of how it was presented and always report “My goodness, this is great stuff!”

Okay, so I'm not as muscular as Jacob, just go with itAnd yes, I labored. I wrestled with my own angels for a while.

And in the end, ET (“Evolution Technology” for newcomers to NextStage’s work) won. Either I accept the tools as accurate or I have no right to expect others to do so. I can’t proclaim “NextStage’s tools are correct for everyone else but me” because that’s simply not how it works. ET is designed to report unbiasedly, to understand human emotion while not being influenced by any emotions of its own (so far. A future release will respond emotionally when asked), so truth is truth is truth and there are no shades of gray in ET’s world.

Gosh, how simple. No wonder it frightens people.

A First Reader Who’s On Her Way to Being the First “Outside the House” Certified NextStageologist suggested I include this Addendum

So the reason that some people stay with NextStage for years and others quickly fold and go away is that scrying part. You may not even want to lift that mirror and being around us, those Principles and all, tends to lift it for you. This scrying isn’t something we do intentionally. It’s probably an aspect of “The Joseph Effect” (see Understanding and Using NextStage’s Level 1 Sentiment Analysis Tool) and just seems to happen.7

I guess this goes in as another undocumented use of ET — scrying.

We have 7+ tools out there now and more on the way. The principals of the company that spawned this post are realizing they have a reflection and congratulations to them.

To whomever else may be reading, “Mirror, anyone?”

1 – I feel another tool coming on…The NextStage Suckometer!

Actually, that wouldn’t be a stretch…According to FireClick for the week I’m writing this (12 Jul 10), global conversions are 1.8% and cart abandonment is 72%. Obviously the sites generating those numbers suck.

People look at NextStage’s KnowledgeShop site and politely let us know it “sucks”. They don’t use that word and it’s in there anyway. And our numbers are…pretty good. I mean, we doubled sales last month. Mostly on bulk purchases, too, not 1-offs, and that includes book sales.

So I’m comfortable with the fact that our site “sucks” and recognize it must suck in a completely different way than other sites suck. I mean, it has to be on a whole different suck system because our numbers are lots better than those listed by FireClick.

Maybe readers should come to us to learn how to make their sites suck, too?


2 – Even when we do “advertise” it’s very dry, statistical, and demonstrably fact oriented in nature, no wild promises, no hype. Exactly what we tell our clients not to do.

But we are NextStage. We follow a different path…


3 – A typical example of the difference between cultures was demonstrated at a meeting many years back and completely unrelated to online analytics (it didn’t exist at the time).

I mentioned that I would no longer be attending the meetings because I was learning more on my own than from the group. One group member challenged me on this, “You think you can sit by yourself and learn more than you can learn here?”

I said, “Buddha sat in front of a wall for days and finally arose, saying ‘Now I’m enlightened’.”

My challenger harrumphed, “So now you think you’re Buddha.”

“No,” I answered. “The wall.”


4 – In fact, way back in 2004, before there were wikis and blogs and MySpace, FaceBook, YouTube, FourSquare and such, we created a site, MirrorOfYourSoul.com (now it points to the Pictou County Flyers site because, between kiting and Nova Scotia, that truly does mirror our souls here at NextStage). That site’s layout was a scrying mirror that allowed users (we tested it pretty extensively with college students) to gather, chat with each other, and offer comments on material that revealed things about themselves.

It was quite the hit and typical to NextStage, having proven the concept, we put it on our shelf and moved on. We really need someone on the “taking the proven concept to market” part, should anybody out there be reading…


5 – I’ve made attempts to get some of the big names in the online analytics industry to look at themselves analytically. Now that, I readily admit, is hard!


We know you were here. We have no idea how much you learned, only you do.6 – Now perhaps you’ll understand why we offer “recognition of attendance” rather than “certifications” in our trainings.


7 – I admit to enjoying learning about The Joseph Effect although I’d rather it be referenced as The NextStage Effect as it’s not about me, it’s about a way of “doing”, of “being”. Example: Susan and I received the following in an email when we returned from a recent research retreat, “I wanted to get back to both you and Susan to say I really enjoyed meeting you both and hope that we can further our friendship. You were kind and thoughtful. Stimulating and challenging. What friends should be. You sent me back to California thinking about how does one live an ethical life without making it a pedantic one:)”

One thing we’ve learned through our studies is in the title of NextStage’s Principles page, “When you squeeze an orange you get orange juice.”, meaning “Apply pressure to a system and you learn how that system really works. If that system is a human, put that human under pressure and you learn what that human is really like, how they really think, whether or not they believe what they claim, can do what they claim, etc.

It’s not about making incredible tools for us, although what it is for us — leaving the planet a better place than we found it, helping people live better lives — is probably what allows us to come up with the tools we do.


Nostra Culpa re NextStage Sentiment Analysis

NextStage’s Evolution Technology calls for human help whenever it encounters something new, unique, or out of its normal experience. Reading Virtual Minds Vol. 1: Science and History readers know our technology does this because I’ve documented it in that book.

This time our system alerted me about a specific Confidence value (from the Intermediate Sentiment Analysis Report) that was a little askew compared to other values it had determined, so I sent an email to the user who’d run the report and offered to go over it with them so we both could learn what that Confidence value applied to.

On Friday (2 Jul 10) afternoon, after our coders had left for the July 4th weekend, the user wrote back very graciously (thanks!) that they’d need to learn how not to fabricate in their writing.


Their response threw me. What did “fabrication” have to do with this Confidence value?

Development History

Readers who’ve followed NextStage Sentiment Analysis development and beta users may remember that NSSA’s Confidence report grew out of a request from FindMeFaster CEO Matt Van Wagner for a tool that could determine if a blog author was full of BlueSky (Matt had another term) or not.

It took a long time to come up with something that I was comfortable with as determining blue sky because there are so many different factors to determining intentional BS from unintentional BS from joking BS from … This discomfort showed up with almost daily rewrites of the Confidence descriptive text. The rewriting process was similar to Mark Twain’s “The difference between the right word and the almost right word is like the difference between lightning and a lightning bug.”

What we came up with was a Confidence equation that included various BS factors because I couldn’t figure out how to completely separate the two (we can discuss the Confidence-BS link at a convention or training sometime, if you’d like. It’s pretty interesting). I wasn’t completely satisfied with the formulation we came up with, could accept it for what it was and told everyone who was using the tool about my concerns.

Then in early May 2010, during conversations with some brilliant researchers specifically about how BS is formed in cognition, we came up with a way to separate BS from Confidence and proceeded to completely spin off Matt’s BS Meter into a separate tool that dealt with whether or not some writing was fabrication or not.

Mea Culpa

But I focus on the charts whenever I look at our reports, not at the descriptive text included in the reports. I’ve been seeing these charts and such for better than ten years at this point so I simply look at the charts, know what’s being reported and respond to that.

I don’t look at the text anymore.

And I obviously should. When this user emailed me that they needed to work on fabrication I went “Huh?” and looked at the report again. “What does this have to do with fabrication?” Then I looked again. Then again. Then I read the report.
The descriptive text for the Confidence report was:

Confidence Gauge – The above gauge indicates (on a scale of -100 to 100) the author’s confidence in their own material. Values from -100 to -80 can most likely be considered pure fabrication although this may not be the author’s intent. Also note that someone writing fiction is intentionally fabricating information. Skilled authors and dramatists can write pure fiction and this meter will indicate confidence is high merely because they have high confidence in their work. This chart is most applicable to people with moderate to no creative writing training.

That descriptive content was the best we could come up with prior to spinning off the BS Meter. The funny thing (to us) was that the suggestions (not shown here) were based on Confidence metrics, had nothing to do with BS and had been part of the Confidence report from the start. Those never changed.

But we’d spun off the BS Meter.

And we’d written new, more accurate descriptive text for the Confidence report:

Confidence Gauge – The above gauge indicates (on a scale of -100 to 100) the author’s confidence in their own material. Some examples:

  • Values from -100 to -75 can occur when the author believes strongly in their material (is confident) and also believes it will not be well accepted, understood or acted upon by their audience (isn’t confident about its reception).
  • Most research and technical writing will score between -20 and 0 because researchers and technical writers tend to have an “I should check this one more time” mindset.
  • It is common for natives of the USA to score between -15 and +10 when analyzing casual, “every day” writing.
  • Truly confident writers will score between 15 and 35.
  • Scores higher than 80 often indicate the author will come off as either sarcastic or vain, based on the author’s acceptance by their audience.

This chart is most applicable to people with moderate to no creative writing training.

And we (I) completely forgot to put it in.

Let this be proof that I’m not as clever as (it seems) many people think.

Making Amends

It’s amusing that this mistake was discovered after we reported our best sales month ever.

But Principles are Principles and when squeezed, one discovers the flavor of the juice.

So by the time this post sees the light of day, everyone who purchased NextStage Sentiment Analysis use since 12 June 2010 (when the BlueSky Meter was released) will have received email notification that their subscription has been renewed. Please contact NextStage if your subscription isn’t renewed (and have your purchase data handy).

Hey, it’s not exactly an oil spill in the Gulf and we do what we can to make things right.

Posted in Analytics, NextStageology, Sentiment Analysis, ToolsTagged , , , , , , , , , ,