Funnel Flow

Landing page: present tripwire product.  Button triggers stripe custom checkout popup.  On form return function(token, args) is called (javascript).  This sets variables in a hidden form (form contains only elements of type “hidden”).  The form’s action is set to a php page that will contain the order bump.  This is submitted once the values (charge token, address, products purchased) are set.

Order Bump:  This page presents the order bump.  It must be php though, because it’s receiving the information from the first page and doing something with it.  This presents 2 options buy or don’t buy.

Buy -> hidden form is updated with one more item, then javascript forwards to upsell php page.  This requires the posted values to move into javascript variables.  To move a php variable in to a javascript variable the syntax is something like:

<?php echo $variable3 ?>

Don’t buy: current form values are sent to receipt.php

Upsell: Buy or don’t buy again, only this time either way goes to receipt

Receipt: This page needs to process the order in stripe, charge the card, send a thankyou email to the customer, and then send a notification email to me that an order needs to be shipped.

So far I have the landing page done, order bump almost done.

Funnel Progress

Funnel progress has been slow.  We had an awful lot to do in the last few days and I didn’t get to it much.  Broadly speaking here are the steps:

  1. Design the landing pages – done
  2. Create a Stripe account to do charges – done
  3.  Figure out how to pass payment information from the form: -done
  4. Figure out how to decrement inventory (will require a database of sales and purchases, this may end up being done through stripe’s orders API
  5.  Add tracking tags and scripts
  6. Go live with the whole shebang!

This required brushing up on PHP and javascript a little unfortunately.  The checkout process works as follows:

  1. Create a checkout form using custom checkout:
  2. Put the values in a hidden form and then post said form to a php script using jquery
  3. php code then will display the next step in the funnel, repeat 2
  4. one of the values is  a “token” that represents the card, the last php script will create the charge, and decrement the appropriate inventories

Exit Intent Forms

Exit intent forms are those annoying popups that come up when you try to leave a page.  I’m familiar with two types:  The kind where you try to close the tab and it opens a bunch of message boxes saying “OMG YOU HAVE SO MUCH VIRUS TEH RUSSIANZ GUNNA PWN U BRUH!”  (or something to that effect), and the ones that ask for your email address to send a coupon code to as soon as your mouse leaves the viewport.

The second type is all the rage right now.  Email address lists are very valuable: a targeted but not very well used or active list is worth $0.20 cents per entry according to that book).  The value of a customer who leaves your landing page leaving no information and doesn’t purchase is close to zero.  It’s not zero, because you may still have retargeting information gathered through tracking pixels, but it’s way lower than the value of an email address.

So how do I do one?  Well much like the funnels people will try to sell you a lot of services.  I don’t want to spend money on these services, since I would like to spend that money on advertising instead.  My helpful link of the day:  This is a quick java script snippet you can use for your own landing pages.

Free Templates

The sales funnel project has several challenges:

  • HTML design of the pages (???):  Unfortunately, clickfunnels promotes so aggressively even finding templates for self hosting is difficult.  I am looking at adapting landing page templates now.
  • Payment processing, inventory, etc. (Stripe API + some custom PHP)
  • Conversion tracking (Mautic)

Payment processing is going to be a bit tricksy but it looks like it’s going to be doable.  Unfortunately I will need to have a separate stripe account for it, which means that I won’t be able to track inventory with shopify.  This is rather unfortunate, but not that big a deal (I have enough inventory.  For now I can just allocate half to each platform and make corrections when sales happen. At some point in the future I can make a shopify plugin that runs on my host that keeps things in sync).

Conversion tracking will happen more or less automatically as long as all the pages are setup as mautic landing pages, and there is a thank you/confirmation landing page after purchase.

HTML templates, as I said are  a big problem, largely due to clickfunnels messing up the signal to noise ratio on this.  I’m going to try to adapt landing pages from

Sales Funnels and eCommerce Evolved

My reading list expanded a little. I got the book “eCommerce Evolved”.  I would recommend it.  It had a great deal of information, and more importantly it wasn’t the same information as other places.  At least some of it wasn’t.  It had hard numbers of good and bad operating margins, conversion rates, etc.  It also had concrete advice of what to automate and how to handle email drip campaigns.  Perhaps most importantly it has step by step market research advice.  It’s by no means perfect but it’s the closest thing to a detailed HOWTO I’ve found yet.

The book suggested reading it cover to cover once, then mulling over the info for a few days, and coming back to it.  When coming back, pick a portion of the book to try to implement in your business, and start there.  I’ve done the first part of that.  I decided, since I have reasonable traffic costs now, to implement the product funnels first.  The book suggests that a physical product funnel will convert several times better than simply pointing an ad to your product page.  In fact it says you should not spend ANY advertising money directing cold traffic to your shopping cart.  That doesn’t include email marketing or retargeting campaigns, those are still OK.

The interesting thing about this is I had been attempting to do similar things with landing pages already, but I wasn’t having much luck.  A proper sales funnel I had not built though.  The reason is mainly the difficulty of it.  I have tried to find a good way to do it with shopify, and the only plugin I found was 99 dollars a month.  That seemed a bit hefty to me for an experiment.  My entire facebook ads budget right now is about 300 a month.

I have two options essentially: Create a chain of static pages that direct to a prefilled shopify cart at the end (easy), or create a chain of pages and use the Stripe API to handle payments.

The book I read was pretty adamant about taking payment first then trying to do upsells and cross-sells, and the data tracking will be a lot easier if everything including the checkout is under my control.  This way I can tag things to be tracked in Mautic (somehow? don’t know how) and use that as triggers for things – customer email list or prospect email list.

Cruise Vacations and Bad News

Well, me the wife and the baby got back from our 1 week cruise in the carribean last weekend.  Aside from the ridiculous hurricane season this year I can only say good things about the trip.  The good people at Carnival are amazing and things run perfectly.  I had a problem with one of the shore excursions, but I’m no worse for the wear.

The bad news, however, is that a good friend of mine was murdered in a mass shooting in Plano Texas.  It started as a domestic dispute seemingly involving depression and alcohol abuse and culminated in 9 lives being lost.

Growtraffic Results

Unfortunately, grow traffic proved out like I suspected.  It is not, at least not obviously, a scam.  It is, however, useless for most purposes.

I purchased 10k visits over 10 days for 28 dollars.  These were directed to a landing page that offers a discount code in exchange for an email signup.  My results ended up being:

  • 4756 hits to my landing page (not sure why it didn’t register 10,000 but whatever
  • 0 email addresses received

Based on that confidence interval calculator I posted a few weeks ago that puts the conversion rate between 0 and 0.1%.  A very similar landing page linked to google ads got 12 out of 100 visitors to give their emails, which gives us 6.6-19.1% conversion.

Conclusion: don’t waste your money.  So far facebook ads work the best, but are also the most difficult to do right.

Common Typos in Searches

Typos are common, and for the most part when searching with google or bing, people don’t even go back to correct them.  Google has been smart enough for a long time to guess what you meant correctly most of the time.  The adwords keywords however are based on exact (but not case sensitive) matches to the search terms.  This is a time for software help.

I haven’t tried this yet but I intend to modify my bing ads campaigns using this data.


Search Ad Keyword Optimization

I read that one thing that you should do to increase CTR on your search ads is try to break them up into lots of customized ads in ad groups, each with only a few keywords.  When someone sees an ad for exactly what they search for they are more likely to click.  For example “red ink pen” should get them “Best value red ink pen, 50% off”.  Not “luxury fountain pens!”.

Naturally this entails writing hundreds and hundreds of ad groups and keyword variations.  But… you can automate this with a perl script or whatever you feel comfortable with.  Simple make templates with tags that can be subbed out.  These then get inserted into a CSV file and reimported.  Both google adwords and bing support this, although for bing you have to use their windows app to import and export campaigns from csv.


So I temporarily got banned by adwords.  My takeaway here Google adwords isn’t the only game in down anymore.  Adwords has since reinstated my account.  But it was very frustrating.  They sent me a notice saying I was hosting malware.  At this point I’m thinking: “are they crazy?”, “did I get hacked”, “are they scanning the wrong site?”.  I called their support line.

The problem is their support people are useless.  You may already know my base domain directs to a shopify site.  I have very little control over that obviously, other than some minor edits to the theme.liquid file that I generally use for tracking codes., however, points at this server, to a mautic install that I’ve been using for email marketing, tracking, and landing page A/B testing.

Naturally my first questions were “which subdomain did you find serving malware”, and “what was the exact URL that you found serving malware”.  They sent me links they had found on my site, but did not answer either of those questions.  They simply kept repeating we gave you all the necessary information, please scan your site.  I asked what scan tool I should use, and no answer.  I eventually found the issue was a typo.  I had writing somewhere through copypasting wrong, was redirecting to one of those parking sites that is all banner ads, and one of those was flagged as high risk.

Now the more infuriating part.  I said I fixed the first link, but could not find the second (I now believe the second was being dynamically inserted by the first “” javascript link).  For a full day, the google representative kept responding that it was still flagging and check again.  I sent him two screen shots of different scanners including google transparency report, and he said those tools aren’t good enough, scan thoroughly.

When I called and asked for the supervisor, that person said my site had been scanning clean since that morning, and should have been cleared for ads already.  He then refused to admit that his employee was lying to me all day.

I will now exclusively use bing ads, they have a better clickthrough rate anyway.