Incoming wall of text:
You need to streamline your production, Concentrate on few income earners and make your production of them as efficient as possible.
You mentioned having farms, make sure you have farmers markets to sell those water melons, they'll be more reliable and make more profit in the long run than relying on B2B.
Stores are more efficient when they sell more types of items, so you should also grow a few more things that will sell in the farmers market.
As far as research is concerned your goal when starting out is to merely get your quality levels close to average, anything above that will be very costly and won't result in significant gains.
I generally raise my research centers to a fairly low size, 10-20 and then queue up research for my chosen item untill the time taken suddenly jumps massively. Then move on to the next item, once your primary income earners have all reached that point you're safe to queue up production of them, then set your R&D to raise the levels higher.
After you start making good profit from produce, you would probably want to build a factory and start producing secondary products, a good choice would be items you can still sell in your initial farmers markets so you don't need to spend on new stores.However you'll notice that fruit plantations do not produce anything that can be converted into other items that will sell in a farmers market, so you could either start building Plantations that produce cereal crops, which you can sell in the farmers market, and also items that can be processed and sell in the markets, or if you have already heavily upgraded your fruit plantations you could decide to build supermarkets, which will not only sell the fruit, but also juices made of them, and ice cream/ sorbets etc.
Another strategy you need to consider is to only upgrade the size of your buildings in ratios that are compatable with your production chain.
For instance my jewellery Business "dosh 4 gold" has 2 size 100 mines, 2 size 100 smelters 1 size 100 jeweller and 3 size 200 stores, and in a 24 hour period the mines produce 5000 gold, the smelters can process 4800 gold into approx 5000 bars, the single jeweller can use 5000 bars to produce 6000 gold bracelets and 2000 gold necklesses, and the stores should sell 6000 bracelets and 2000 necklesses.
Everything produces almost exactly the right quantities to make things efficient, so when upgrading i need to take that into account, and expand in roder of the production chain, for instance 50% increase in mine capacity, then 50% more smelter, 50% more jeweller then 50% more stores.
The last thing you need to organise to be efficient is the sales price in your stores, 2x is usually far from the most efficient sales price, if you have a huge stockpile of something that you really want to get rid of you may lower the price to barely above what you paid just so it moves, and if you picked something up at a very low price you might go much higher than 2x to maximise profit.
The general rule i use is to modify sales price so that i sell slightly less every day than i produce.
So for instance those 2000 necklesses would amount to 20.83 per tick (2000/(24*4)) so i adjust sales price untill the estimate per tick is approx 20, then i watch it for a few ticks to see if it really sells that well.
Then when you expand store capacity you'll find that the sales volume goes up, so you can raise sales price to which drops volume but raises profits massively.
Also when you expand and your store capacity goes down temporarily you could lower price to keep sales volume high, or just ignore it and let your inventory go up a little.
Using this system you'll find that you sometimes end up with a little excess inventory, which you can sell off on b2b for quick profit every so often, or you could micromanage constantly to raise production/sales temporarily in order to offload excess(which is what i do but is rather time consuming).
And finally you might want to demolish some buildings that aren't helping you make profit, i know those size 100 research centers are cool and cost a damn lot to start with, but are basically useless if you can't actually afford to research at the rate they provide.
The building upkeep on them alone will eventually cost more than the construction cost, and you get a significant refund for demolishing buildings. If your profit is so bad and your loan so high it may be the cash injection you sorely need to streamline yourself back into the black.