how i went from zero to 22.93% conversion rate on twitter ads in 1 wk.
How i went from zero to 22.93% conversion rate on Twitter ads.
Twitter is great. It dominates the real time space and that can make it a very effective test bed for marketeers in the social space.
Twitter also provides very powerful tools in-built with the Twitter advanced search feature.
Twitter is also very powerful for the quick and cheap advert test with very highly targeted advertising.
However, twitter is NOT like any of the other social media and can easily be abused.
Twitter ads are a form of content marketing on their own and need to offer value for the reader as well as harmonize with their interests.
How I went from 0.0% to 22.93% conversion rate on Twitter ads.
There were a total of 7 tests performed from Aug 12th onward.
The best conversion rate of the first 6 was 2.47% which was actually pretty good as I am now informed. Our cost per conversion on our best test was at .31 each.
Our 7th test results were achieved at a cost of .03 each! Yes, 3 cents each. At that cost, cheap to experiment and tune the results!
The 7th test I used the software from my predictive linguistics work, the 'emotive reduction engine', and ran the twitter-feed through it for about a half an hour, isolated by an advanced search targeting specific words. I then used the results to tune my LP and my tweet. Then I launched the campaign and went to over 20% within the first few minutes. I knew I had the formula at that point. Conversions were also analyzed against follow-on behavior by visits to an amazon store I had set up for the test.
Test number 7 was profitable within a few hours.
The emotive reduction engine approach harmonizes with the underlying emotional tones of the tweetees such that we arrive at this initial level of the mid 20's in conversions.
Our next goal will be for a 50+ % conversion off Twitter ads.
Come and see what else I do with the emotive reduction engine at....
For more information on the current and past ALTA future forecasts please visit
http://www.halfpasthuman.com/#alta_sales_area
http://www.halfpasthuman.com/#timetalks for the latest 'timetalks' discussion about all things time...
http://www.halfpasthuman.com/#chemies for How to Breathe Free in a Chemtrail World! This was the LP site for the twitter advert campaign! As you may imagine, not an easy sell at any level, so well chosen to be my LP for this experiment. Our conversion rate was here, at this level in SPITE of the nature of the content, and entirely due to the emotional qualifiers/quantifiers in our lexicon. The attractant line on the Twitter advert is
How to Breathe Free in a Chemtrail World!
Follow on twitter at @clif_high
subscribe for latest up youtube videos
#clif_high
#webbothits
#webbot
#webbotreport
#halfpasthuman
#chemtrails
#chemtrails2017
#socialmarketing
clif high | web bots | predictive linguistics | future forecasts | halfpasthuman | ALTA reports | web bot reports | trends | web bot hits |social media marketing | twitter ads | twitter marketing| social marketing |
So stepping outside of my usual line of work here, which is all future-based and focused on making forecast using predictive linguistics, which is basically a science I invented in 1993 and have been working on ever since to step into the world of social network marketing.
Not a social being, I wouldn't know how to network if I dripped over them and don't do marketing with crap, which is why I'm poor.
So but the good news about this is that I know how to research.
I'm not stupid, and I can teach myself so I can educate myself around all of those, and I've gotten some pretty good results so far, and thus this quick little video about how I went from very little to uh 22% greater than a 22% conversion rate on Twitter ads.
And it wasn't that hard.
I only did seven experiments and it didn't cost hardly any money at all.
So quick uh bit of background because I'm going to be drawing in some um uh non-woo uh people uh from the social marketing realm uh due to how this will be labeled, uh bit of info about myself.
My name is Cliff High.
You can look me up, C L I F-I-G-H.
I uh invented predictive linguistics, uh I'm a software engineer and a programmer.
I wrote a lot of code that I call the uh emotive reduction engine that runs my predictive linguistics forecasting.
I make uh predictions of the future that have had a pretty good hit record, especially recently, and uh put these out on a site called halfpasthuman.com.
You can go there and check it out.
You can also uh follow me on Twitter at uh Cliff at uh at Cliff Underbar H I G H. Uh only one F in Cliff for real-time updates that I drop out on uh Twitter, and that's uh pertinent.
We'll get to that in a second.
Uh the subject of this uh video is how I went from zero to 22 plus percent conversion rate on Twitter ads in only seven easy steps, and you don't even have to do the first six.
So, anyway, I mean to get to this.
Um continuing about that real quick though, uh uh my emotive reduction engine was a great deal of um uh software engineering that I did around converting words into emotions, or rather how to extract emotion out of the language uh such that you can store a representation of it in the computer and program against it to do other things with words.
And we'll let it go with that.
Let's get into our subject right at the moment about Twitter and uh Twitter ads.
Uh Twitter is great.
Uh they say there's uh been a billion people that have gone through Twitter and 400 million of them have stuck.
Uh what's uh perhaps even more interesting is 283 million, something like that.
Uh people go through Twitter and are active on it in any given uh week.
Um Twitter is an interesting vehicle for myself because I deal in future and time and that sort of thing.
Uh Twitter is really interesting because it's real time or as near real time as you can get in a large aggregated uh uh database uh matching algorithm.
And uh it's uh very interesting that way because it's real time, marketers can use it to do incredible uh searches such as I've done, which led to my achieving this 22% conversion rate on this Twitter ad.
And also, by the way, the conversion rate was at three cents a hit.
Uh staggering.
It's like whoa, um not only 22, not only one greater than one out of five uh people seeing the ad went right to my link and jumped into that engagement with me on Twitter, but it was only costing me three cents a pop to get them to see the ad to begin with.
And this was only my seventh ad experiment.
I like Twitter because it's real time, because it's real time, you can put together a little structure quickly.
Uh I did my entire testing campaign for Twitter in about four hours.
I even built a back-end Amazon store as a secondary uh tracking mechanism for um follow-through for actual commitment to the uh uh conversion by spending money.
All of this took as I say about four hours to construct the web website, do the landing page, uh, do the back-end Amazon store, and do the advanced searches for the Twitter uh campaign.
Now the Twitter ad campaigns that I did, I did the first six based on information that I found on the uh net for from um social marketeers who were telling me, you know, basically how to structure large-scale SQL queries because underneath all of these engines is structured query language, and use those queries to get vast quantities of people and try and match them up against interests and so forth.
I did this on my first six campaigns.
My best hit record on my first six campaigns was 2.47% conversion rate.
And the cheapest one I'd gotten out of those six campaigns was about $0.31 per conversion.
So again, very expensive conversions.
Some of them went as high as 98 cents.
And it was a question of how I was doing it relative to the supposed quality of the conversion targets.
And I say suppose it because there's a tendency to put value that may not be pertinent to you.
We look recently at Procter and Gamble, for instance, having left Facebook because they were not getting value for their money.
And so in my first six campaigns, I was not getting value for my money as an advertiser.
And this is my first go-round on advertising.
I've never really done this before.
And so on the seventh campaign, I decided, well, what I would do is to ignore the standard marketing mindset towards reaching into clients.
Even though they were basically aiming towards a content marketing, as I was, and uh express it in a different way, and that is to allow my work to find a self-selecting audience based on the words in which I was uh interested in uh using as this vehicle.
So I ran my predictive linguistics uh spider algorithms against the Twitter feed in real time again, maybe 28 minutes, uh something like that to get a good base, and uh tagged off a bunch of emotive words around certain uh keywords, uh tagged those to tweets and investigated the context manually.
That is to say I went and looked at the tweets individually after they were highlighted by my software, aggregated into a collection, and then I took those collections and made some subtle changes to the lexicon I normally use for my predictive linguistics and changed it into sort of like uh ad conversion uh linguistics.
And this um uh then I went back to the content used for uh that's backed by the Amazon store and um uh adjusted the wording, the preferences, the word order and the uh uh statement types and some of the grammar to reflect the uh stated intent so that we have integrity, right?
Because your your uh your thoughts, actions, and um words all have to be tight in order to have integrity, and uh you've got to be integrity to what you're putting out there.
And so uh I went back and did that and adjusted my uh landing page for that, adjusted the content, um, came back to Twitter and uh ran an advanced search to come up with a slightly finer degree of uh keywords, and then I built some demographics that were would shock you because I was dealing with uh 16 keywords, uh two behaviors, one area, and uh two interests, I believe.
I'll have to go back and double check that.
And that uh was my seventh test, and within uh you know, Twitter's real time, it was near real time.
They have to digest your query, get it all set up, translate it into SQL, and then fire it off, wait for it to fill into tempdB a bit, and then start spitting the results back to you.
And all of that took probably about maybe a couple of minutes.
I went I went out and and made myself a little bit of tea while it was happening.
Came back and uh watched the results spit out, and the results were very impressive.
Almost instantly I was about uh 12 to 15% conversion rate.
And I have, as I say, the Amazon back-end store that tracks the commitment to the conversion, plus I have my own demographics and logs and analytics on the web page.
So and then because I was tracking for engagements, I was also able to track it on Twitter relative to the analytics that they had there.
Relatively simple process, didn't cause me too much mind meltdown.
A lot of marketing speak usually does because I don't understand it.
But at this level, I certainly do, obviously, to go from 2.47% up to over 22%.
And it actually reached 22.93% that evening before I went to bed.
And that's when I stopped worrying about the numbers.
I achieved what I was after with that particular test.
And so to do that, obviously the predictive analytics approach works, where I allow the minds to be self-selecting, is basically what I'm doing.
I find subjects that it that the mind is interested in.
I don't care so much about the body that's in that the mind is associated with at that stage or what's in the pockets of that body.
That's kind of immaterial.
But we all have to admit that the point of the conversions is to allow for the expression of energy, which frequently comes in the form of uh purchases, cash, donations, or whatever, in order to be able to tie into that, which is what my uh point was with this.
We have to have our conversions provide value for the individuals that are being converted.
I mean, it's a quid pro quo in universe.
Uh, you get value for the value that you offer.
And so my um landing page for this was a very good landing page.
It provided uh a lot of information that these people had never seen before.
Uh and it was a um uh success as well because I went was able to instantly go back, not instantly, I think it was like six hours later.
I was able to get some information out of uh Amazon and their uh analytics.
Indeed, we had had a uh successful number of sales relative to the uh conversion rate, and in fact, had made uh greater amount of um uh fees or whatever they call them, uh they're not commissions, they're um affiliate fees or whatever Amazon calls them.
We'd made uh slightly more than uh twice the cost of the Twitter ad to that point.
So at that point that I was uh doing the analysis on the Twitter ad, we'd spent 43.96 cents.
We'd converted 1200 uh 47 people, something like that, at a cost of uh according to Twitter, we were paying three cents uh per conversion, and the Amazon store had taken in about eighty dollars at that stage.
It was like 79 or 78, something like that, and so it was running ahead of the cost of the advert.
So a very uh profitable uh little test there.
Now I've shut that down and I'm working with my um lexicon to retune this such that I can hopefully achieve a better than a 50% conversion rate and retarget for a wider number of audiences and for a different kind of um landing page and a different level of conversion tracking uh on my own.
So uh it's a very interesting process to um for us anyway, and so far small but profitable test to abandon uh the usual way of thinking about marketing and allow our customers to be self-selecting around the language and the emotional component of that language and the objects and the uh context in which those objects are being referenced.
Uh not trying to be uh obtuse or oblique here, it's just that this predictive uh linguistics, the emotive reduction engine is complicated, it reduces uh emotions uh that are tied up in words down to uh quantifiers and qualifiers, and then we make some assumptions based on those.
So it it get it does get a little bit um technical.
Uh and uh but at this stage it it's uh also a little bit profitable.
So this is cool.
Uh I'll keep you guys updated as I go forward with this because it's quite fascinating, and there I have other um motives for releasing uh videos on the social marketing part of it, obviously.
Um experiments uh took about a week.
I started on the 12th uh learning about this stuff, uh read up about how the um the 12th of August read up about how uh Twitter and so forth works.
It's fascinating, fascinating.
Uh and also uh much maligned and and not well used by most of the people that are marketing on it.
They just spam it, they just don't understand what it's all about.
And they're really better off on the uh on the other more static uh social media.
Uh but Twitter is quite fascinating with its real-time dominance, and for marketers that uh have an inquiring mind, uh real time is where it's at.
So this was uh basically a little video on um how I got from zero to uh 22% conversion rate in only seven small ad experiments with Twitter ads.
And you can follow me uh at Cliff Underbar, C-L-I-F, underbar H-I-G-H on Twitter.
I uh dump my real-time um updates from our predictive linguistics on Twitter when I know they're gonna be stale by the time I can get them into a report.
It's the easiest way to put it in there.
Check us out.
Uh you can look us under up under Twitter for Webbot Hits.
