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. mautic.myshopifydomain.com, 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 .com.com somewhere through copypasting wrong, com.com was redirecting to one of those parking sites that is all banner ads, and one of those was flagged as high risk.
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.
I mentioned yesterday that I was intending on making a landing page and then using grow traffic to direct traffic there. I suspect this is going to be a pointless waste of money, since I read rather abysmal reviews. But – right now spending about 30 dollars on google ads only netted me 9 subscribers That’s like 3.50 per email subscribtion, so I don’t need a high conversion rate to improve on that.
Anyway, I cloned a landing page in mautic and was about to get going with this experiment. And I noticed my form was giving 403 forbidden instead of redirecting to my shopping cart. I still haven’t really figured out why. For now the work around was simply to clone the form too. But I will need to play with this and see if I can duplicate the error reliably. Preferably with pages I’m not using.
So I’m needing to get a little more flexible in driving traffic to my squeeze pages. I found this which claims to use a combination of methods (unfortunately including popunders).
Is it a scam? Does it work? Will it just piss everyone off? Lets find out! I purchased the minimum order (10000 visitors over 10 days for 28 dollars). That’s cheap enough to risk it. I’m directing it to a mailing list sign up page. I’ll post results in 2 weeks assuming it actually goes live as promised.
So I signed up for bing ads. I think I’ll give a quick opinion. The basic interface is essentially a straight up knockoff of adwords. That’s convenient, since it makes moving campaigns between the two easy, and eliminates the learning curve. It offers an option to import your campaign directly from your adwords account as well. That’s convenient.
Now the negatives: The payment interface is lame, it recognized my credit card wrong (entered as mastercard, kept showing up as AmEx) and I wasn’t sure if it was working. It turned out it’s just a bug in the system and wasn’t affecting my payments but still. After importing the campaign the ads kept showing up as “your account or campaign is paused”, but it turns out they were just not showing yet. Other platforms generally say something like “pending review” so you know you just have to be patient. This combined with the payment thing made me think something was wrong with my account.
And more positives: Click through rates and CPC are slightly better at least for my ads. I was getting terrible click through rates on adwords, in some cases around .1%. Yes I could tweek and tweek and optimize and get those up. But they went up to about .5% for the same ads and keywords on bing. I have to assume it was better placement due to less competition. I don’t care where my traffic comes from, and I have no love for google so this is fine by me.
I was playing with my mautic landing pages on my laptop and I noticed the form was hanging at “please wait” instead of redirecting. It appears that the redirect URL must be hosted the same as the form. Since shopify requires all sites be served using https, and my server was not requiring this, I had logged into mautic over insecure http on my laptop. It took my a while to figure out that was what was making the form hang. The simplest solution to this is to modify your virtual hosts file to redirect http requests to the https site.
You can follow these instructions: https://www.namecheap.com/support/knowledgebase/article.aspx/9821/38/redirect-to-https-on-apache
In the line that is
Redirect permanent / https://www.yourdomain.com/
do not forget the ending “/” or else the redirect will fail
I typoed that the first time and redirects were failing, what’s worse my browser caching was preventing changes from seeming to take effect. It’s a good idea to test your pages in a new anonymous/incognito whatever your browser calls it window so you know caches and cookies and such aren’t affecting anything.
A/B testing, or split testing is the worst kept secret of the big players in tech. Basically you randomly serve your visitors one version or another of a page. You then define a metric for success, such as click throughs, subscriptions, or sales. Then you test which one does better. Switch to serving only that. You can rinse and repeat this and evolve your marketing material into the most effective form.
There are lots of tools for split testing, but I’m using the landing page split testing built into mautic. For beginners, especially those of us on a shoestring budget, it’s important to be aware of statistics. We typically have low conversion rates, and so we need to be able to calculate when we have a significant result.
I’m recommending this tool: https://www.peakconversion.com/2012/02/ab-split-test-graphical-calculator/
We’ll look at the math behind it at some point, but basically the bayesian approach is more sound than most significance testing out there. There are lots of approaches to calculating confidence intervals. A lot of them break down pretty badly when you have low probability events.
I promised musings so here are some musings
On advice from people: Don’t talk to friends and family about trying to start your business too much. If you do make sure you are talking to the right ones who will be encouraging but realistic. Red flags are people who routinely say things like “oh yeah I should do that that would be so easy”, or “that can’t be so simple there must be a trick”. These statements seem like opposites, but they are actually the same thing. The person is rationalizing why you are wrong to be even trying, and they were right to settle for wage slavery while their telomeres inexorably shorten leading them closer to the abyss.
In the first case they are just so gosh dern busy but if they only tried they know they would be way good at it with no effort at all! This person needs to think they are more capable, and also more successful in other areas of life than you. They are so busy! They have a great social life and friends and family that’s why they don’t have time for all that stuff you do! It’s not laziness, no sirree bob, you are the loser. They will subtly discourage you, latch on to any failures as justification for not trying.
In the second case the person thinks there is some trick, some formula only an elite class have access too. Perhaps something that one can purchase from a business school (you can’t by the way, MBAs are for Corporate America, not kitchen table startups). That is their justification for remaining subservient to their masters. This person will encourage you to follow the status quo of office politics and die of a heart attack shortly after finally getting the big promotion to middle management. Your wife and child will get a letter from your boss about how much they appreciate all the hard work and that you gave your life to increase diversynergy on the deep dive team of six sigmafied agile belts. They will also get 1.5 years of income as a life insurance payout, and receive a bill from the emergency room for 3x that amount.
OK so I have a few updates. First of all I made some changes to the virtual hosts file on my server to allow for additional domains. I’ll put up a link to a how to on that. The important thing if you are using it for mautic is to edit the .htaccess file in your new webroot for your virtual host. I had problems because it was not redirecting correctly. It’s a one line fix, check your htaccess and see if there is a line still pointing to the wrong directory.
Secondly, I did this for a couple things, now I have a subdomain of my shopify domain that points to mautic on my server. That means that I will be able to link to mautic landing pages and it will not appear to be leaving the domain of the site. That’s good because redirects to offsite look fishy.
Thirdly, I installed a second instance of mautic, and linked it to this site. That way I will be able to use this site as a sand box for things like popup forms, pages, etc, before going live on the real ecommerce site.
First email results are in from Mautic:
That’s about 30% read. That’s much higher than I expected. With such a small list I wasn’t expecting to get sales every email so I’m happy with these results right now.
My first email with mautic didn’t go as planned. Once again waking up I looked at the results and nothing had been sent. The problem this time however was of a technical nature.
Setting up Cron Jobs
My crontab now looks roughly like this. You will need to change the paths to match your mautic install.
* * * * * php /var/www/html/mautic/app/console mautic:segments:update
*/4 * * * * php /var/www/html/mautic/app/console mautic:campaigns:rebuild
*/5 * * * * php /var/www/html/mautic/app/console mautic:campaigns:trigger
0,30 * * * * php /var/www/html/mautic/app/console mautic:emails:send
15 5 * * 5 /opt/letsencrypt/letsencrypt-auto renew >> /var/log/le-renew.log
These cron jobs are all necessary for mautic to do it’s job. The first syncs email addresses in all the segments, rerunning the filters. The second updates any campaigns, and the third, which I was was missing, triggers events in campaigns. The fourth is only necessary if you queue emails. You may want to do that if you have outgoing mail restrictions on your hosting, but linode does not. I put it there anyway for future proofing.
The last job in my crontab will auto renew my lets encrypt certs. To add these to your cron jobs just run the command crontab -e. It will ask you to select an editor (vim ftw but you can use nano if you want).
Go ahead and copy paste one of those commands into a terminal and make sure it runs. In my case it didn’t, since I didn’t have permission to write everything in that directory. Everything in there belonged to both group and user www-data (the user php executes as). This may have been some mistake I made during installation, but nonetheless if you run across it here is what I did:
sudo chmod -R g+w mautic
sudo usermod -aG www-data breandan