Delivered is not the same as in the inbox
Inbox placement is the question of where your email ends up after the receiving server has accepted it: in the inbox, in the Promotions tab or in the spam folder. Delivery only measures that acceptance. A campaign can therefore be one hundred per cent "delivered" and still go largely unseen, because the recipient does not tell you which folder the message landed in.
What "delivered" actually means
When your sending server offers a message, the receiving server answers "accepted" or returns an error code. That answer is the only hard fact you get. What the provider does next it decides on its own, based on your reputation, your authentication and the behaviour of its users, and it does not share that decision with you. A delivery rate of 98 per cent therefore only says that 98 per cent of your messages made it through the front door. Whether they then ended up in the living room or the cellar is not part of the figure.
Why this misleads
The classic situation: the newsletter reports a tidy delivery rate and a strikingly low click rate. The marketing team gets to work on subject lines and buttons, while the real problem is that a large share of readers never saw the message at all. They are optimising the content of a message that is sitting in the spam folder.
How to recognise a placement problem
You cannot see the folder directly, but you can infer it. Three signals point to it:
- A drop at one provider. Compare opens and clicks per receiving provider. If you see ten per cent opens at Gmail addresses and thirty at the other providers, while delivery is the same everywhere, Gmail is filtering you out. A content problem would hit every provider equally hard.
- A sudden break over time. Opens that halve from one send to the next without any change to your content point to a change in reputation, not to a less interesting subject.
- Test addresses that confirm it. A handful of your own addresses at the large providers, included in every send, show you literally which folder the message falls into. It is not a statistic, but it is the fastest proof.
What usually fixes it
- Get authentication in order. Without SPF, DKIM and DMARC the provider decides against you by default. See SPF, DKIM and DMARC.
- Remove inactive addresses. This is counter-intuitive but it works: less unread mail means a better ratio, and that ratio is exactly what the provider uses to place your next send. Anyone who opened nothing for a year no longer belongs in the main list. How to do that without losing customers is explained under list segmentation.
- Spread volume more evenly. A spike of fifty thousand messages in one morning after weeks of silence reads like an attack. Regularity is rewarded.
- A real unsubscribe button. Anyone who cannot unsubscribe easily clicks "spam". Every complaint pushes down the placement of all your following messages.
What does not fix it
Cutting words from your subject line to dodge a supposed list of "forbidden terms". That list does not exist the way people imagine. Filters look at who you are and how your recipients react, far more than at what you write. Announcing a price cut with the word "discount" in it is no problem for a sender with a good reputation, and a spotless text does not save a sender with a bad one.
And the Promotions tab?
That is not spam. It is a sorting that some providers apply to commercial mail, and many readers look there deliberately. A newsletter in Promotions is normal; an order confirmation in Promotions is a sign that your transactional and commercial mail look too much alike or travel the same route. Separate them, both in content and in sending channel.
How to follow it
You follow placement not per message but per send and per provider, over time. For that you need figures per receiving domain: delivery, opens, clicks, bounces and complaints side by side. The deliverability page explains which figures the dashboard shows per domain and how to spot a drop at a single provider quickly.
Frequently asked questions
Can my sending platform see which folder my mail landed in?
No, no platform can do that directly. Providers do not share that information. What a platform can do is put the figures per provider side by side so the deviation becomes visible.
How do I measure placement without open tracking?
With test addresses at the large providers that go out with every send, and with the number of replies and clicks on your links. If you do not measure opens you cannot see the folder, but you can see whether people respond.
How quickly does placement recover after cleaning the list?
Count on a few sends. The provider needs to see new, better signals before it revises its decision. One clean send is not enough; three in a row usually is.