Spare Parts Planning and the Service Network in India

Brands plan finished goods well. Most of them plan spare parts badly, and then they are surprised when a service network they paid for cannot actually service anything. The reason is that spares are a different inventory problem wearing the same clothes, and the planning method that works for a fast moving SKU actively fails on a part that moves twice a year.

Why spares are not finished goods

Three things separate them. Demand is a long tail. A durable product with two hundred components will see meaningful failure in maybe fifteen of them, and the rest sit as insurance. Velocity is low and lumpy, so the demand history is mostly zeroes with occasional spikes, which is exactly the pattern that breaks a standard moving average forecast. And the service impact of a stockout is wildly disproportionate to the value of the part.

That third point is the one that changes the maths. A missing forty rupee gasket does not cost you forty rupees. It costs you a customer waiting three weeks, a technician visit that has to be repeated, an escalation on a marketplace, and often a full unit replacement because the desk gave up waiting. The cost of the stockout is many multiples of the cost of the part, which means the usual service level logic based on part value is upside down.

Forecasting also has to change. For slow, intermittent demand, do not use a simple average. Install base logic works better. Take the number of units in the field, the age profile of that install base, and the observed failure rate of the component, and derive expected annual demand from that. A part fitted on a hundred thousand units in the field with a one percent annual failure rate needs a very different plan from a part fitted on eight hundred units.

Criticality classification comes before any forecast

You cannot stock everything everywhere. So classify first. Two axes are enough for most Indian brands.

The first axis is functional criticality. Does the part failing stop the product working entirely, degrade it partially, or affect only appearance. A compressor is a stopper. A knob is cosmetic. Both are spares. They do not deserve the same treatment.

The second axis is supply risk. Is it an off the shelf item you can source in a day from a local vendor, is it a proprietary moulded part with a single tool, or is it imported with a long lead time and a minimum order quantity. A proprietary imported part with a twelve week lead time has to be planned months in advance regardless of how rarely it fails.

Cross those two axes and you get a simple grid. Stoppers with high supply risk get deep, forward positioned stock and a hard minimum. Stoppers with low supply risk can run lean because you can replenish quickly. Cosmetic parts with high supply risk get a decision, either buy once at tooling stage or accept that you will not offer them. Cosmetic parts with low supply risk get ordered on demand.

Add one more input that brands routinely forget. End of life planning. When you discontinue a model, the install base does not disappear. Decide at discontinuation how many years you intend to support it, and place a final buy accordingly, because the tool will be pulled or the vendor will move on and there is no second chance.

Where to hold the stock

The instinct is to push spares out to every service point. It is the wrong instinct for most brands. Spares behave like slow moving inventory, and slow moving inventory pooled centrally needs far less total stock than the same inventory scattered across twenty locations. Scatter it and you will simultaneously have a stockout in Kochi and dead stock in Ludhiana.

The workable structure for India is usually three tiers. A central spares hub holds the full range, including everything slow and expensive. Regional nodes, typically aligned to your existing warehouse footprint, hold the fast moving and critical subset for the states they serve. Technician boot stock holds only the parts that fail often enough to be predictable, the consumables and the top failure items, because a technician arriving without the part means a second visit and a second visit destroys your economics.

The number that decides tiering is first visit fix rate. If your technicians resolve on the first visit less than about three quarters of the time, your boot stock list is wrong or too short, and the fix is almost always in the top ten failure parts rather than in broadening the range. Broadening boot stock across the long tail is expensive and does not move the number.

Set replenishment for spares on a review cycle, not on a reorder point alone. Weekly review at regional nodes, monthly at technician level, with automatic top up against a fixed maximum. Simple, visible, and it survives staff turnover.

Service partner or in house

There is no universally right answer, but there is a right sequence. In house first, in your top demand cities, while the product is young and the failure modes are not yet understood. You need your own people close to the failures in the first year because that is where the design feedback comes from, and a third party will not give you that detail.

Partner networks make sense once the failure modes are known and documented, and once the repair can be reduced to a procedure. The trap is that brands hand over the network before they have written the procedure, and then blame the partner for a quality problem that is really a documentation problem.

If you partner, control four things contractually. Parts must come from you or from an approved source, because counterfeit parts inside your warranty are your liability. Turnaround time has to be measured by you, from the customer’s clock, not from the partner’s ticket. The claim data has to flow back to you at serial level. And rates should reward first visit fix rather than pure visit volume, otherwise you are paying a partner to visit twice.

Watch the coverage map honestly. Partner networks look national on a slide and are thin in practice outside the top forty cities. If your demand is moving into smaller towns, your service map has to move first, not after.

Spares availability is a repeat purchase lever

Most brands treat spares as a cost of doing business. They are actually one of the cheapest retention levers available in a durable category.

The logic is direct. A customer whose product failed and was fixed quickly is measurably more likely to buy from you again than a customer whose product never failed, because the fix is the only time your brand did something visible for them. A customer whose product failed and could not be fixed becomes a permanent detractor and takes their category recommendation with them, which in Indian households is worth several purchases.

There is a second commercial effect that gets ignored. Published parts availability supports resale value, and resale value supports the price a new buyer is willing to pay. It is the same reason a two wheeler with easily available parts holds price. If you also run a refurbished channel, spares availability is the input that makes the whole channel viable.

So measure the things that reflect this. Parts fill rate at the point of the technician, not at the central warehouse. First visit fix rate. Age of open service tickets, watching the tail. And repeat purchase rate split by whether a customer has ever raised a service request. That last cut is uncomfortable and it is the one that will get spares budget approved.