Buy a trial quantity first, in almost every case. A first fresh food order answers two questions, whether your dog will eat it and how fast your household gets through it, and neither needs a large purchase. Move to a full order once the recipe is proven and the portion is confirmed.
The first order has a job, and it is not feeding
Treat your first purchase as data collection. You are buying answers to four questions, and the quantity you need is the smallest one that answers them.
Will your dog eat this recipe, consistently, past the novelty of the first two days? Does it agree with them, judged by stool quality and by whether anything changes in their skin or energy? How much do they actually need, measured rather than estimated? And how fast does your household physically get through a delivery, including the days someone forgets to move food from the freezer?
A large first order does not answer any of those any better than a small one. It just raises the cost of a wrong answer.
When a full order is the better first purchase
There are three honest exceptions, and it is worth stating them because the trial first advice is not universal.
If you have fed this exact recipe before, perhaps at a previous address or for another dog in the same house, you already have the answers and a trial is a delay.
If you have a large dog and a trial quantity would last only a few days, the trial may be too short to tell you anything. A few days covers the novelty period and nothing else. In that case a moderate order is more informative than a small one, though still not the largest available.
If shipping is a significant part of the cost and a trial quantity pushes you below a free delivery threshold, the cost per day of the trial can be much higher than the cost per day of a full order. That does not make the trial wrong, but it does mean you should price it honestly rather than assuming small equals cheap.
Outside those three, start small.
Read the last two rows carefully. A trial is excellent at answering the dog questions and poor at answering the household questions. That is the honest limit of the approach, and it is why the sequence matters: trial to settle the recipe, then one full order to settle the logistics, then a cadence.
How to run the trial so it produces a usable answer
Transition gradually over roughly a week, increasing the proportion of new food and decreasing the old. Feed at the same times, in the same bowl, in the same place, and add nothing.
Weigh your dog at the start of the trial and again at the end. This is the single most useful thing you will do, because it converts the guideline portion into a real one. Without it you will carry an estimated portion into a full order and your cost per day will be wrong from the first delivery.
Keep a note, even a rough one, of how many days the trial quantity lasted and how much was left when you reordered. That is your consumption rate, and it is what you will use to choose a delivery cadence later.
Do not judge on the first two meals. Novelty produces both enthusiasm and hesitation, and neither predicts what happens in week two. Portion accuracy and body condition are the things to watch across the whole period, and the American Veterinary Medical Association's pet owner resources are a reasonable plain language reference for what good body condition looks like while you are assessing.
Moving from trial to full order
Once the trial has produced clear answers, place one full order before you commit to a recurring schedule. That order tells you the two things the trial could not: whether your freezer holds a real delivery, and what the cost per day looks like once shipping is spread across a proper quantity.
Then set a cadence based on your measured consumption rate rather than on the brand's estimate for an average dog. Review it twice a year, because dogs change weight and activity with age and season.
If the trial went well and you are choosing a recipe to scale up with, a single protein option keeps the reasoning simple. The Pork Blend is a useful example of that shape, with one named protein rather than several, so if anything changes later you have fewer variables to unpick.
One practical note on figures. The numeric panel, meaning calorie content, guaranteed analysis and feeding directions, is not printed anywhere on a My Perfect Pet product page. Use the weight based calculator on the page for your starting portion, and get the printed numbers from the pack, the Nutrition Facts Sheet PDF for the recipe, or a written reply from the brand. Do not build a full order around a portion you estimated.
Frequently asked questions
Should I buy a trial size of fresh dog food first?
In almost every case, yes. A first order exists to answer whether your dog will eat the food, whether it agrees with them, and roughly how much they need. A small quantity answers all three at a fraction of the cost of finding out you were wrong.
How long does a fresh dog food trial need to be?
Long enough to cover a gradual transition of about a week plus at least another week of feeding the new food alone. Anything shorter only measures novelty. For a large dog, check that the trial quantity actually covers that period before assuming it will.
Is a trial size more expensive per day than a full order?
Usually yes, because shipping is spread across less food and small quantities rarely reach free delivery thresholds. Price the trial as a one off cost of information rather than as a representative cost per day, and recalculate once you place a full order.
When should I switch from trial orders to a subscription?
After one full order, not straight after the trial. The trial settles the recipe and the portion; the full order settles whether your freezer holds a real delivery and what the true cost per day is. Set a cadence only once you have measured your consumption rate.
What should I record during the trial?
Your dog's weight at the start and the end, how promptly they eat at each meal, stool quality across the second week, and how many days the quantity lasted. Those four notes convert a trial into a portion, a cadence and a budget you can rely on.
Related reading: Subscription or One Time Order: Which Fresh Food Buying Model Fits Your Household, and How to Pick a Starter Recipe When You Do Not Know Your Dog's Preferences.