If you do not know your dog's preferences, start with a variety format or a familiar single protein, avoid anything they have reacted to, and change one thing at a time. Pick the recipe closest to what your dog already eats, then let two weeks of real feeding decide. Preference is discovered, not predicted.
Work from what your dog already eats
You know more than you think. Look at the bag or can your dog is on now and read the first three ingredients. That is your baseline, and it is the best available predictor of what will go down easily.
If the current food is chicken based and your dog eats it without fuss, a chicken based fresh recipe is the lowest risk starting point. The protein is familiar, the gut is used to it, and if your dog refuses it you have learned something specific: the issue is the format or the texture, not the protein.
If the current food is a blend of several proteins, you have less information, because you cannot tell which of them your dog likes. In that case a single protein start is more informative, even if it is not obviously more appealing.
Note anything your dog has reacted to in the past, whether that was itching, loose stools or vomiting. Exclude those proteins from the starting shortlist entirely rather than choosing a recipe that merely contains less of them.
The four inputs that narrow it down fast
Age and life stage come first. A puppy needs a recipe formulated for growth or for all life stages, and that requirement overrides preference completely. An adult dog with no particular needs has the widest choice. A senior with a reduced appetite often does better with a stronger smelling recipe, which in practice often means fish or a red meat.
Known sensitivities come second, and they are exclusions rather than preferences. Rule out, do not rule in.
Texture and aroma habits come third and are underrated. A dog who has only ever eaten dry food may hesitate at a soft texture for the first few meals. That is a novelty response, not a rejection, and it usually resolves within three or four meals if you do not intervene by adding things.
Your own constraints come fourth. If one recipe costs meaningfully more than another and both are reasonable starting points, start with the cheaper one. You are gathering information, not making a permanent commitment, and information collected cheaply is information you are more likely to collect.
Run the first two weeks as an experiment
Change one variable. That is the whole method, and it is where most people go wrong, because they switch food, add a topper and change the feeding time in the same week and then cannot interpret the result.
Transition gradually over about a week, mixing increasing amounts of the new food into the old. Feed at the same times, in the same place, with nothing added. Resist the urge to help a hesitant dog by stirring in something tasty, because that teaches waiting rather than eating and it contaminates your result.
Watch three things. Does your dog eat it within a few minutes, most days? Are stools firm and consistent by the end of the second week? Is body condition steady? Those three answers tell you whether the recipe works far more reliably than the enthusiasm of the first meal, which is often just novelty.
Weigh your dog before you begin and again at the two week mark. The American Kennel Club's nutrition guidance is a good plain language reference for what to look for in body condition while you do this, and the habit will serve you long after the starter recipe question is settled.
If the result is ambiguous, extend to three weeks rather than switching again. Switching twice in quick succession makes the next result harder to read, not easier.
Using a variety format to shorten the process
When you genuinely have no information, buying several recipes at once is more efficient than buying one and guessing. You get direct comparisons in the same dog, in the same week, under the same conditions, which is the only kind of comparison worth much.
The Sampler Pack is built for this. It lets you observe which proteins get finished first and which get picked at, and it also gives you a real measurement of how fast your household gets through a delivery, which you will need later if you set up a regular order.
One caution: a variety format is a discovery tool, not a diagnostic one. If you are trying to identify a food sensitivity, rotating proteins actively works against you, because several variables move at once. That job needs a structured elimination plan agreed with your vet.
The product pages at My Perfect Pet carry no ingredient declaration, no calorie content and no feeding directions, so take a starting portion from the weight based calculator on the page, and get the printed figures from the pack, the Nutrition Facts Sheet PDF or the brand in writing.
Frequently asked questions
How do I choose a first fresh dog food recipe?
Start from the protein your dog already eats without trouble, exclude anything they have reacted to, and match the life stage requirement. If your dog currently eats a blend and you cannot tell which protein suits, a single protein recipe or a variety pack will give you a clearer answer within two weeks.
Should I buy a variety pack for a picky dog?
Usually yes, when you have no information about preferences. It produces several direct comparisons from one purchase, under identical conditions. It is the wrong tool if you are investigating a suspected food sensitivity, because rotating proteins moves several variables at once.
How long should I feed a new recipe before deciding it works?
Two weeks after the transition is complete, and three if the result is unclear. Judge it on whether your dog eats promptly most days, whether stools are firm and consistent, and whether body condition is steady. The first meal tells you about novelty, not about fit.
My dog sniffed the new food and walked away. Should I switch?
Not yet. A dog moving from dry food to a soft texture often hesitates for the first few meals. Keep serving it at the same times with nothing added, and give it three or four meals before concluding anything. Adding toppers to tempt them teaches waiting rather than eating.
Does the starting recipe matter if I plan to rotate later?
It matters for the first few weeks, because a clean start gives you a baseline to compare against. Begin with one recipe, confirm it agrees with your dog, then introduce rotation gradually. Starting with rotation leaves you with no reference point if something goes wrong.
Related reading: Trial Size or Full Order: How to Buy Fresh Food the First Time, and Single Protein or Blended Recipe: Which to Buy First.