Planning With Registry Demand
A baby registry is one of the few genuinely forward-looking demand signals in consumer retail — created months before it converts, itemised, and explicit about intent. This guide covers how to bring registry data into a merchandise plan, the conversion lag, where the signal misleads, and why it usually sits in a channel report instead of the forecast.
A rare kind of signal
Almost every demand signal in retail is a projection of the past. A forecast extrapolates history, demand sensing reads recent movement more quickly, and both are ultimately statements about how well the future will resemble what already happened. That is why forecasting gets hard exactly when it matters most — at launches, in new categories, and after a structural change.
A baby registry is a different kind of object. It has three properties at once that are rare individually and nearly unique in combination:
- It is created well before it converts. The list exists months ahead of the purchases it will generate.
- It is itemised. Not a category interest or a browsing signal, but named products.
- It states intent explicitly. The customer has said what they intend to acquire, rather than leaving it to be inferred from behaviour.
That combination means a juvenile brand has access to something most categories simply do not have: a forward read on demand that has not happened yet. The baby and juvenile planning guide treats this as one of the defining features of the category. This guide is about actually using it.
Why it usually goes unused
Before the method, the obstacle — because it is not analytical.
Registry data typically arrives through a retail partner's reporting or a DTC channel team. It is therefore owned by whoever owns that channel, and it is presented in the form that channel reports in: performance, after the fact, in a monthly or weekly review. It is looked at. It is discussed. And it is not in the plan.
Meanwhile the buy is committed on a different cadence, by a different group, working from a forecast built on shipment and sell-through history. The registry signal and the decision it could inform never meet. Nobody decided to ignore it; the organisation simply has no mechanism for a channel report to become a forecast input.
The fix is structural rather than motivational: carry registry as a named demand input against the same plan as sell-through and shipments, so it participates in the forecast rather than commenting on it afterwards.
Measuring the lag from your own data
The conversion lag — the time between a registry add and the purchase it eventually produces — is the parameter that makes the signal usable. Two disciplines matter.
Measure it, do not assume it. Registries are generally created well ahead of the birth and convert in a concentrated period around showers and the final weeks before the due date. That general shape is stable enough to plan against, but the specific distribution differs by brand, by retail partner, and by the mix of registrants. It is measurable from your own history and should be.
Measure it separately by item class. This is where blended analysis quietly fails. A travel system is a high-consideration purchase, frequently bought by the parents themselves or by a group contribution, often late, and sometimes after substitution. A pack of bibs is a low-value gift bought early and impulsively. A single blended lag describes neither, and using one will systematically mistime the buy for both — early on the big-ticket items where capital is tied up, late on the consumables where availability drives conversion.
A workable segmentation is by price band and consideration level rather than by merchandising category, because the behaviour tracks the size of the decision more than the department.
Where the signal misleads
Registry data is strong but it is not an order book, and treating it as one produces confident errors. Three failure modes are worth naming.
Adds are aspirational. Add-to-purchase conversion is well below one and varies systematically by price point — the more expensive the item, the more likely it is aspirational at the moment of adding. A plan that treats adds as committed demand will over-buy at the top of the range, which is precisely where over-buying is most expensive.
The purchaser is usually not the registrant. Gift-givers select from the list, and they select according to their own budget and their own relationship to the parents. Completion discounts then shift some of the residual back to the registrant late in the cycle. This changes both timing and mix: gift-givers cluster on mid-price items, while the expensive items disproportionately fall to the completion window.
An unconverted add is not lost demand. It is very often substitution within the same class — a different stroller, not no stroller. So item-level registry data overstates certainty about the specific SKU while understating the strength of the category signal. This is the single most important nuance for how the data should be used.
Using it at the right level
Those failure modes point to a clear method: plan the envelope at category level, use item share as an input to mix.
At category level, registry volume and composition give a genuinely reliable forward read. The number of active registries containing a travel system, weighted by the measured conversion rate and offset by the measured lag, is a defensible input into how much of that category will sell and when. This is where the signal is strongest and where it should carry real weight in the buy.
At item level, registry share is directional. It is useful for ranking within a class, for spotting an item gaining or losing share against its siblings, and as one input among several into the depth decision. It is not a commitment, and the substitution behaviour above is the reason.
Used this way, registry data does something specific and valuable: it improves the timing and size of the category buy, which is the decision with the longest lead time and the most capital at stake, while leaving item-level depth to the usual sell-through evidence.
RetailNorthstar has no baby gear or juvenile products customers today — apparel is the flagship vertical and that is where its track record is. What it offers here is the ability to carry multiple demand inputs against one plan, so a forward signal like registry volume participates in the forecast alongside sell-through and shipments rather than living in a separate channel report. The registry platforms themselves, and the retail partner reporting that feeds them, sit outside the system.
What it changes about the buy
Bringing registry demand into the plan changes three decisions in particular, and it is worth being concrete about which.
Category buy timing. Because the signal leads conversion by a measurable interval, it lands early enough to influence an order that has an ocean lead time in front of it. This is the highest-value use, since it is the decision that is otherwise made furthest from the demand it serves — the point the baby and juvenile planning guide makes about cube and container economics applies directly here.
Run-out forecasting. During a model-year changeover, the question is whether remaining inventory will clear before a fixed boundary. Registry volume against the outgoing platform is a forward read on exactly that, and it is available earlier than sell-through can show the same thing.
Channel mix. Registry behaviour differs between retail partners and DTC in composition and in conversion. Reading them separately, rather than as one aggregate, informs where depth should sit — which matters more in a category where a single door's minimum is a large physical commitment.
The connected version
None of this requires exotic analytics. It requires the signal to be in the same place as the decision.
In a connected model, registry volume is a demand input against the same assortment plan that carries platform and colourway structure, feeding the same buy plan that sizes orders against lead time and container economics. The lag and conversion assumptions are parameters of the plan rather than of a monthly analysis, which means when they are re-measured the forecast moves with them.
The alternative — the one most brands are living with — is a good signal, correctly analysed, arriving in a report that nobody can act on by the time the buy is placed. For the wider category picture, see the baby and juvenile industry page.
See how RetailNorthstar carries a forward demand signal alongside sell-through in one plan.
Book a Demo →Related resources
- Merchandise Planning for Baby & Juvenile Brands — The pillar this guide sits under
- Planning a Model-Year Changeover — Where registry data improves run-out forecasting
- Baby & Juvenile Brands — RetailNorthstar — Platform fit for juvenile planning teams
- Demand Forecasting for Fashion — How forecasting works when no forward signal exists
- Sell-Through Rate Formula — The evidence base for item-level depth
- Demand Sensing — Glossary — Reading recent movement faster, and why that is a different thing
- Demand Planning — Glossary — Where a forward input belongs in the process
Common questions
What is registry demand?
Registry demand is the demand signal created when expectant parents add items to a baby registry — a list built months before the items are purchased, itemised at product level, and explicit about intent. It differs from most retail demand signals in that it is genuinely forward-looking rather than a projection of past behaviour, and the intent is stated by the customer rather than inferred from browsing or purchase history.
How is registry data different from a sales forecast?
A sales forecast projects the past forward and is therefore only as good as the stability of the pattern it extrapolates. Registry data is an observation about the future: the customer has already told you what they intend to acquire, and roughly when the need arises. The practical difference is that a registry gives a read on demand that has not happened yet, whereas a forecast gives a read on demand that resembles demand that already happened.
How long is the lag between a registry add and a purchase?
It varies by item class and by who buys, so it should be measured from your own data rather than assumed. Registries are typically created well before the birth and convert in a concentrated period around showers and the weeks before the due date, which makes the lag long enough to inform a buy and stable enough to be useful. The important discipline is to measure the lag separately for high-consideration items like travel systems and for low-value consumables — they behave differently and a blended lag describes neither.
Where does registry data mislead?
In three predictable places. Registry adds are aspirational, so add-to-purchase conversion is well below one and varies by price point. Completion discounts and gifting behaviour mean the purchaser is often not the registrant, which changes both timing and item mix. And an unconverted registry is not lost demand — it frequently converts to a different item in the same class, so treating the specific SKU as committed overstates item-level certainty while understating category-level demand.
Why do most brands leave registry data out of the plan?
For organisational rather than analytical reasons. Registry data usually arrives through a retail partner or a DTC channel team and is reported as channel performance, so it lives in a channel report and is presented after the fact. The buy is committed elsewhere, on a different cadence, by people who see the report but have no mechanism to carry it into the forecast. The signal is not missing — it is simply not connected to the moment where it could change a decision.
Can registry demand be used at item level or only at category level?
Both, but with different confidence. At category level it is a strong signal: the number of active registries and their composition gives a reliable read on how much of a category will be bought and roughly when. At item level it is directional, because registrants substitute within a class before purchase and gift-givers substitute again. The workable approach is to plan the category envelope from registry volume and use item-level registry share as one input into the mix decision rather than as a committed order book.
Share this guide with your team
Copy a link or a pre-written message for Slack, Teams, or email.
// Know where your operation stands
Apply this to your planning operation.
The free Apparel Planning Maturity Assessment benchmarks your operation and tells you exactly which gaps to fix first.