List Cleaning - Why Isn't There an Easy Way?

Alex Read

Active Member
Hi

Everywhere I read, it's important to have active & engaged subscribers to get a higher inbox rate.

What's the current best way to remove 'inactive' subscribers?

Maybe have 2 fields:
  1. Sent Count
  2. Open Count

and if Sent Count = 5 AND Open Count = 1, send them a final email and move to an 'Inactive' list?

Or is there a better/easier way?

Concerns:
  • When I check open count, sometimes a subscriber can open it once, but because they use a mailclient, it get's counted like 5X, so not sure this is a reliable number we can use.
 
Great to read this today as I'm dealing with exactly this same scenario (once more).
I have many subscribers with double-digit open-counts just 30mins after sending a campaign. Too many to explain this through forwarding or other means. The tracking pixel must be loading on a loop or through a proxy on the webmail side before adding tracking to all subscribers using that same mail client.

This could be related to my ongoing bugbear: the device stats: https://forum.mailwizz.com/threads/devices-statistics.1306/#post-44599
I'm slightly disappointed that this issue has not really been looked at (or maybe it has but it's not being talked about more)

Regarding segmentation, I'm struggling to even get the simplest 3-way rule running on a mini-list of <3K records. I make exactly the same query as you, a co-relation between engagement (in my case a combination of opens and clicks) and sends (in order to keep new signups included). I have now also tried adding the "number of campaign opens" over time - but that's already too much to even go through and gives me a "too deep" error:

Screenshot 2021-02-11 095244.jpg

So I'm a little disheartened with regards to the segmentation capabilities. I'm on a fast server and if this is something I can ease on my side, it would be great.
 
You can give it more time, by default we only allow 5 seconds to load the segments, but give it < 30 under Backend > Settings > Customers > Lists > "Max. segment wait timeout" (same if this customer is in a group) and that should do it.
If you feel things are slow, increase server resources.
 
When does a subscriber falls in the category to be cleaned?
Just my 2 cents: That's completely dependent on contact frequency. Hence my recommendation to give control to the list owner via a decent segmentation engine. MailChimp does this via their star system (which is kind-of a dumbed-down reflection of the data). Others have different methods.

It all points to the power of the options within the segmentation in MW, a first step having been the recent implementation of "number of openes over time". Yet it's not quite fit for automation since we need to be able to keep new subscribers included as @Alex Read pointed out. Plus a decent way of combining ANY with ALL queries. In addition, a further metric beyond opens are clicks to measure engagement.

That's why I started applying a points system for each open and click, utilising the [INCREMENT_BY_X] dynamic field content... only to realise that this gets screwed by the messed up open-stats (back to the Google image caching). It's all interconnected.

Sorry about the brain dump.
 
That's why I started applying a points system for each open and click, utilising the [INCREMENT_BY_X] dynamic field content... only to realise that this gets screwed by the messed up open-stats (back to the Google image caching). It's all interconnected.
Exactly this.

@twisted1919 - you can't use open count as it's not accurate at all (when I checked a year ago).

How would you go about removing disengaged subscribers?

See here https://optinmonster.com/email-scrubbing-how-to-clean-email-list/
You can choose subscribers who did not open:

  • All of the last 5 campaigns
  • Any of the last 5 campaigns
  • All recent campaigns
or
Inactive (or cold) subscribers are the subscribers on your list who haven’t opened any of your emails for 90 days.

You see, these are 'global' checks on an account that would apply across all lists. I.e. if a subscriber hasn't opened any of the last 5 emails, update their status to 'NOT ACTIVE' and then we can target them in another list.

Something like that....
 
You see, these are 'global' checks on an account that would apply across all lists. I.e. if a subscriber hasn't opened any of the last 5 emails, update their status to 'NOT ACTIVE' and then we can target them in another list.
We're almost on the same page, @Alex Read. But this particular setting should be optional.

I have targeted lists and some users are highly engaged with particular lists and not so with others, depending on the content/topic/subject. I wouldn't like engagement metrics to ever be global. That defeats the point of letting good content lead to high engagement.

However, yes, the reporting capabilities must be improved first. That's a priority.
 
I just wanted to share this extremely poignant article from AlchemyWorx which outlines exactly what we've been talking about and shine a light on the importance of Open Reach:
Worth a read to understand how important the synergy between open rate, send count and broadcast frequency is.
 
This is a very important subject.

Is there an easy way to identify and delete any subscribers who have not opened any email from any list in the past 30 days?

I'm not interested in identify inactive in each list... I want to know which subscribers, in my entire customer account, have not opened any emails at all for 30 days.
 
@twisted1919 I thought I'd use the filter feature to identify and delete subscribers who don't open in a 30-day period.

However, when I try to use the subscriber filter feature (filter those who did not open any of the campaigns in the last 30 days), the results are horribly inaccurate.

I ran this filter on my "test" list that contains my 21 seed addresses. I ran it several times, using increasingly more filters, and the results were always the same... it showed that 2 of my seed addresses had not opened any of my emails in the last 30 days. However.. that was WRONG. I kept careful records, and 8 of my seed addresses had not opened an email in the last 30 days.

Is there some way to fix this so that we get accurate results and can use the filter feature to keep our lists clean?
 
This is on everyone's mind and you are right. But we cannot tackle this until we have a solution for overcoming the inaccuracy due to Google's image proxy server. Other platforms don't have this problem, so there must be a way around it. I'm sure @twisted1919 and team are on the case.
But I agree, it's absolutely vital. I'm currently checking how Mautic is handling this.
 
I found a workaround. It's not perfect and it's not automated... but it will do until this issue can be fixed.

I created a custom tag for each list that tracks number of opens. Then, for each campaign, I'll have to create a segment that populates that tag if someone opens.

Once a month, I'll have to download each list, identify all the subscribers who didn't open... and then delete them from the spreadsheet and upload the remaining subscribers to a new list.

If anyone has a suggestion for an easier way, I'm listening. :)
 
How do you deal with brand new subscribers who haven't had a chance yet to open an email yet?
Also, why delete? Why not simply suppress them or create a dynamic active segment that updates itself automatically with "opens". That way you don't even have to touch the application and just let it run. All you need to make sure is to only send to the active segment.

As we explained a few times now, the "open" event is not yet 100% accurate yet in MW, because of externally served images i.e. by Google/Gmail. The first Gmail user who opens one of your campaigns skews the tracking count and stats for all other Gmail users, who will thereafter trigger the tracking image on the proxy server, not download a fresh one.
 
@nadworks thank you for explaining in more layman's terms what is causing the inaccurate open events.

What you're suggesting makes sense, except for one thing (and maybe I'm misunderstanding what you're saying). If I set MW to copy or move the subscribers to another list upon open, and then only email to my active lists, that will work.

What I don't understand is how the inactive subscribers might become active if I don't mail to them. And, if I'm not mailing to them, wouldn't it make sense to remove them from the database?

Based on what you wrote, it makes sense to just leave the inactive subscribers alone (at least until MW has figured a workaround). However, I have a very large - and growing - database... I don't want to overwhelm it with inactive subscribers.

If you have anything additional to share, let me know... I'm very open to your suggestions. I truly appreciate your input.

P.S. I have decided to not look for another interim solution, and to just find a way to move forward with MW. I'm very impressed with the software and with their support... just a few little bugs need to be worked out (and I can be patient for that)
 
What you're suggesting makes sense, except for one thing (and maybe I'm misunderstanding what you're saying). If I set MW to copy or move the subscribers to another list upon open, and then only email to my active lists, that will work.

What I don't understand is how the inactive subscribers might become active if I don't mail to them. And, if I'm not mailing to them, wouldn't it make sense to remove them from the database?

Based on what you wrote, it makes sense to just leave the inactive subscribers alone (at least until MW has figured a workaround). However, I have a very large - and growing - database... I don't want to overwhelm it with inactive subscribers.

In my experience, opening emails is organic and not always linear. Some subscribers might miss the odd email you send but open the next one. If you immediately remove people who didn't open the first email that comes along, you're not giving them a chance to engage upon a 2nd or 3rd try. So, build rolling time into your suppression strategy.

If you only suppress records, rather than delete them, you can try to reactivate them with a re-engagement campaign later. I'm assuming that the emails you're sending are all going to genuine subscribers, who have given you contact permission. If you're spamming, applying a spray and pray approach or if you are contacting people without their consent, I'm not keen on even engaging with you. However, if your records are all genuine people who have requested to hear from you or your organisation, then why manage your database with a sledgehammer?
 
@nadworks I've been marketing online (using email) since 1998 (no, I'm not a spammer, lol). I've found that, if someone doesn't open your email in the first 30 days, they're unlikely to open anything you send... in fact, it usually indicates they used a fake email address to sign up for your lead magnet (although who knows why since, if they use a fake email, they never get the lead magnet... go figure).

Anyway, when you include a bunch of non-openers in your mailings, your mailing reputation suffers and your emails start to go to the promotions tab or worse. That is why it is so important to keep a clean list.

I have people joining my list every day. I have no need to waste my time (and hurt my ability to reach the people who want to receive my emails) by mailing to people who don't open my messages.

Before I delete the non-openers, I send them a short series of messages in my last-ditch attempt to get them to open. If they still don't open, then I delete them.

Since I know I'll never bother emailing them again once they ignore those final messages (why hurt my emailing reputation and destroy my ability to reach the inbox?) it just makes sense to delete them from the system. With all the new subscribers coming in daily, I'd rather keep only active subscribers on my list.

I may not be a techie, but I DO know email marketing quite well. So, while there's nothing wrong with your approach, I prefer deleting them from the database to make room for others who want to hear from me.
:)
 
Back
Top