| A multi-faceted project, this was one part:- |
|
| Hold enough to serve their 'little and often' business. |
| Get (and return) their large, infrequent (and non-stock) items from (and to) the centre. |
| Where small orders stop and large ones begin is a matter of judgment every
bit as much as science, and we should learn to trust our judgement. |
| The browser allowed us to combine science and judgment (and to cut through a situation which was out of control) |
| But the best way to run it from now is to use more judgment and only occasional science. |
|
| A: We don't, because that's the wrong way to look at it:- |
| Whether little and often orders coincide, whether future demand mirrors past demand, whether large orders come in when stock is close to max or right on min, is all a game of chance. |
| In a game of chance our aim is to be 'nearly right a lot of the time', and to handle the exceptions well. |
| There's no such thing as 'right'. And if there were we should be backing horses not hiring Supply Chain Tools! |
| The thermometers of whether we're nearly right include:- |
| Does it feel right? |
| Are the annual no of TOSs per item no per branch 'close enough' to what we thought they would be? |
| Is there a review and feedback process to:- |
| Promote and relegate products to and from branch stocked |
| Review the branch minima both routinely and when there's some other reason to do so (e.g. customer or sales feedback,
branch space, stock shortage) |
| Are those processes right, and being followed? |
| If 2 people do the task independently, do they come up with wildly differing results. |
| If so, alarm! |
| If they are in broad agreement it's likely ok. |
| If in exact agreement, look out. |
| They are either collaborating, or using the same rule of thumb. |
| Yet we want them to use judgment, and an ingrained rule of thumb displaces thinking. |
| Does it matter? (If small cheap items are overstocked, do we care?) |
|
| Some will have sufficient history (to begin with they may have been served from central stock). |
| Graph the hire diary and set the minimum stock to cover all the little and often orders and none of the order spikes. As a guide that's probably about 3 central TOSs (Stock transfers) a year. |
| Some will have no history, in which case set a minimum based on what you think the little and often demand will be. |
| If you think the new product substitutes an old one, adopt the old Min. |
| If you think it will behave 'a bit like Product 1234' then use 1234's min. |
| You can always tune it later |
| That sets the min. The max is 2 x min, plus a bundle quantity. |
| First time round the mins were set higher than strictly necessary where we already had excess stock. That, and the fact that returns often keep the stock well above the minimum, means we can shave the mins when we want to. |
| To be blunt, 3 TOS a year per branch per stocked product is a guideline. |
| We would barely notice any difference if some moved twice and others 4 times. |
| The art is 'being nearly right much of the time'. |
| There's broad similarity when setting up a new branch. Take an intelligent punt and tune it later. However, a word of warning. This gives the starting stock, not the ultimate stock (and thereby yard size). |
| Apart from site compounds we shouldn't be building branches unless they can store everything their area will eventually need. |
| And there's relaively little difference in the 'best practice' storage area required for a busy and less busy branch. It's the marshalling area which changes, by far more than the storage area. |
|
| The min stocks can and should be tuned |
| One way to tune them is to track the TOSs and lower the mins on those that |
| Have fewer than 3 TOS a year and |
| Still have some spiky orders. [If all the orders are little and often, it's OK to cover them from branch stock] |
| Are expensive &/or bulky |
| Conversely, consider raising min stock if there are many small TOSs |
| Another way is to use intelligent rounding. If 2 x min + a bundle gives (for example) 3 TOSs a year and requires we store 168 items. Suppose that's one stack (6 high x 25 per bundle) and 8 loose items. Life would be simpler all round, and overall costs lower, if we simply left the min alone and altered the max to 150. |
|
| There's a shortcut way to setting new min / max when combining branches. |
| If the combined total number of hires is less than about 10-15 a year, simply take the higher of Branch A and Branch B stock |
| If it's higher than this, here's the trick |
| Take min for branch A. I'll call it MinA. Do the same for MinB |
| Square the mins and add them together (MinA x MinA) + (MinB x MinB) |
| Take the square root of the total. That's the new Min, MinCombined |
| The max for the combined branch is 2 x MinCombined + a bundle |
| Either way, round it down to a full stow, full stack or manageable quantity if your judgment says you can and should. |
|
